Attachment 8 to Northern Gateway Reply Evidence

Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills

Enbridge Northern Gateway Project

Prepared for: Northern Gateway Pipelines Limited Partnership

Prepared by: Stantec Consulting Ltd. AMEC Earth and Environment Chumis Cultural Resource Services and Coastal, Assessment, Liaison, and Monitoring

July 2012

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Northern Gateway Project Executive Summary

Executive Summary

In written and oral evidence before the Northern Gateway Joint Review Panel, a common statement is that spills from the proposed pipelines and tankers calling on the proposed Kitimat Terminal are inevitable. It has also been asserted that the effects of an – particularly in a marine environment – would be permanent. An example of the arguments advanced to the effect that ecosystems and their Valued Ecosystem Components (VECs) do not recover from spills can be found in submissions filed on behalf of the United Fishermen and Allied Workers Union. Also, many laypersons have stated similar opinions concerning the biophysical environment, either directly to the Joint Review Panel, or in the course of written comments. In written and oral evidence before the Joint Review Panel (JRP) for the Enbridge Northern Gateway Pipelines Project (NGP), some interveners have focused on the Valdez oil spill (EVOS) as their “model of choice” for assessing potential direct, indirect and cumulative effects on the human environment, including effects on Aboriginal people and coastal communities (including non- Aboriginal people). These interveners also claim that a catastrophic marine oil spill is inevitable, and that such a spill will result in permanent social, cultural and ecological damage to coastal First Nations and coastal communities and their way of life.

Purpose and Scope The purpose of this evidence is to reply to the assertion that ecosystems do not recover from spill events. Recovery in both biophysical and human environments is addressed. Using a case study approach, scientific literature regarding past spill events is examined to document recovery of ecosystems from events that have occurred in environments similar to that of the study area (e.g. Cold Temperate or Subarctic, rather than Subtropical or Tropical). It will be seen that although oil spills have adverse effects on biophysical and human environments, the scientific literature is clear that ecosystems and their components recover. The literature is also clear that although recovery can occur in some circumstances without clean-up, appropriate clean-up does enable and accelerate recovery. This review also discusses measures such as bioremediation and phytomediation that advance recovery. The geographic focus of the recovery review is on the north central coast of for marine aspects and on the northern area of British Columbia and for the pipeline aspects. Because of the similarities between the marine areas of the north central coast of British Columbia and the southern part of , as well as the large extent and duration of studies, specific attention is given to the (EVOS). Similarly, the Pine River oil spill is referenced due to the proximity of this spill to the proposed pipeline right-of-way for Northern Gateway. The review provides only brief descriptions of the effects of the specific spills considered and provides references for further spill details. The main text emphasizes the course of recovery. While recognizing that clean-up response is commonly but not always the first step to environmental recovery, the various case studies considered in this review typically focus on the course of events after the clean-up phase has ended and examines the processes and outcomes of

July 2012 Page i

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Executive Summary

the biophysical and human environments that lead to recovery. Prior to presenting specific information on case studies and recovery of biophysical and human environments, the review first describes the general behaviour of oil in the environment. Definitions of recovery are also discussed.

Behaviour of Oil in the Environment As soon as oil is spilled into the terrestrial, freshwater, and marine environments numerous chemical, physical and biological processes immediately begin to break down, biodegrade and otherwise assimilate the spilled oil. This natural degradation of oil sets the conditions under which the recovery of the biophysical and human environments from oil spills occurs. Ultimately, spilled oil is broken down into carbon dioxide and water by sunlight (photolysis) and microbes (biodegradation). Degradation rates depend on the oil type and characteristics of the receiving environment, such as temperature, sunlight, and prevailing microbial populations.

Defining Recovery Definitions of ecological (biophysical) recovery have both regulatory and scientific contexts, and in both, the definitions have changed substantially over the last several decades. The common element in most definitions is a return of the ecosystem or a particular valued ecosystem component (VEC) to some desirable system state following a disturbance. This review examines how recovery is defined and measured. Recently, recovery has been defined as a return to the conditions that would have prevailed had the oil spill not occurred. This definition recognizes the need to account for natural variability and for the influence of natural and man-made factors other than the spill. Recent scientific literature indicates that the goal of recovery and of active restoration of damaged environments should be to recover or restore the ecosystem to a functional state that provides valuable ecological goods and services. This perspective is also emerging in the regulatory context in North America.

Approach Two basic approaches were employed in preparing this review. First, the scientific literature in peer reviewed journals and government reports was reviewed to identify and acquire information on recovery from oil spills. Second, from the oil spills where recovery was followed, case studies of the recovery of specific species, groups of species, human uses and human values were selected using a set of a priori criteria. Because the review focuses on recovery rather than on effects, the selected case studies need to have been studied sufficiently well to assess whether recovery had occurred. For each oil spill and each Valued Ecosystem Component (VEC; for example fish and fish habitat) that were examined as candidate studies, several questions were asked: • What VEC was affected? • What was the reported effect?

Page ii July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Executive Summary

• Was recovery reported? • If the resource recovered, what processes were involved? • Were there obstacles or complicating factors that impeded recovery? • Were there attempts to intervene to accelerate recovery? • How long did recovery take? The results of this process are tabulated in Appendices A and B. From this set of candidate oil spills, specific spills and VECs were selected for more detailed treatment as case studies. Assessment techniques were noted where they were clearly explained. Generally, the review and appendix gives the state of recovery reported by the study authors. For many spills, the study ended before full recovery was achieved. Such studies were scored as recovering and the study duration since the time of the spill was listed. Partial recovery was listed where the study authors indicated that only partial recovery had occurred.

Synopsis of Results The main question in this review is whether ecosystems and the human environments associated with them do or do not recover from oil spills; the answer is that recovery occurs. The scientific literature reviewed here is for cold temperate and sub-arctic regions and for marine, terrestrial, and freshwater environments. The bulk of the evidence is that for marine, freshwater and terrestrial oil spills, recovery of both the biophysical and human environments does occur. The literature clearly reveals that recovery is not a rare occurrence; in fact, recovery is more common than not (Sections 5 and 6, Appendices A and B). For biophysical environments, there were 140 valued ecosystem components for which there were sufficient data for study authors to draw conclusions about recovery (Appendices A and B, Figure 7.1). The VECs associated with marine environments were 69% of this total, and 86% of these marine VECs were recovered or recovering at study’s end. VECs associated with the freshwater environment were 16% of the total and 70% of the freshwater VECs were recovered or recovering. The terrestrial VECs were 11% of the total and 50% of the terrestrial VECs were recovered or recovering. Of the total VECs examined from all environments, 80% were recovered or recovering. Where VECs had not recovered, this often reflected the study duration, or complex interactions with other natural or man-caused factors (e.g., killer whale, herring). These results are consistent with the findings of other review of recovery of ecosystems from disturbance. Such reviews indicate that for such disturbances as overfishing, logging, mining, eutrophication, and invasive species, recovery frequencies of indicator variables were less than 50%. Recovery does take time; how much time depends on the environment, the VEC, and other factors. The average time to recovered status for the biophysical VECs examined here was 2.3 years for freshwater environments and 5.8 years for marine. The frequency of recovered freshwater VECs is highest for cases less than one year and few freshwater VECs have a time to recovery greater than two years. The frequency of recovered marine VECs is highest (33%)

July 2012 Page iii

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Executive Summary

between 5 and 10 years, and about 8% of the marine VECs have recovery times greater than 20 years. The recovery times for terrestrial VECs are more complicated. Terrestrial VECs that recover appear to do so in about 2 years, but in some circumstances time to recovery for terrestrial VECs can take long time periods. Several of the case studies examined in this review were intentional experiments to study recovery or ones with little or no effective cleanup. Commonly, cleanup for terrestrial spills involves the removal of oiled soil and vegetation, replacement with clean soil, and replanting. Where clean activities have not been applied or applied ineffectively, some terrestrial VECs that were examined in this review were still recovering after 20 years (i.e., vegetation and soils in arctic regions) or showed no evidence of recovery (i.e., a contaminated groundwater aquifer).Other reviews of recovery show similar times to recovery, e.g. within 2-10 years for marine spills. A review of 51 cases studies of recovery of coastal ecosystems from a variety of disturbances concluded that recovery can occur in less than 5 years, but recovery from long-term insults, such as mine tailings and chronic wastewater discharges, may take decades. Another study of a number of types of disturbances and ecosystems found that time to recovery from oil spills was less than 5 years while time to recovery from overfishing was 10 to 20 years. Recovery of forested systems from logging and deforestation takes 3 to 4 decades, essentially the length of the cycle to regrow trees. A number of studies have demonstrated that recovery from marine spills is a function of the degree to which the system is sheltered from physical oceanographic processes. At one extreme are open marine waters that recover rapidly in weeks or months, and at the other are sheltered, soft-sediment marshes that recover slowly over two or more decades. In the middle are headlands and exposed rocky shores that take 1 to 4 years to recover. Because of the long recovery times for sheltered systems, modern spill response gives high priority to preventing oil from entering marshes and other similar systems. The same physical, chemical, and biological processes (e.g., spreading, dispersion, evaporation, biodegradation) that govern the behavior of oil in the marine environment also govern oil behavior in the freshwater environment. Where freshwater systems are fast flowing streams and rivers, turbulence can disperse oil into the water column and enhance dissolution and evaporation. Where freshwater systems are slow flowing or relatively still, oil can accumulate and persist longer. Clean-up is generally a positive factor in ecological recovery from oil spills but it can be a negative in some instances. The exceptions to the recovery time of 1 to 4 years for rocky shores were associated with bull-dozing of the shore in the Esso Bernicia and the use of first generation dispersants in the Torrey Canyon. During the Cadiz, removal of oiled sediment from salt marshes with heavy earth-moving equipment lowered the level of the marshes and changed the patterns of sediment deposition and conditions for the growth of marsh plants. Modern spill response includes procedures to carefully select the most appropriate treatment for the oil type, level of contamination, and the nature of the shoreline (e.g., exposed and rocky versus sheltered and soft sediment).

Page iv July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Executive Summary

The characteristics of the spilled oil and the receiving environment interact to determine the environmental fate of the oil and its persistence in the environment. Appropriate clean-up is a positive factor in ecological recovery by decreasing the level and duration of the oil in the environment. A consistent theme emerging from the review identifies two factors related to the life history of the organisms involved. First, recovery proceeds more rapidly when there is an abundant supply of propagules close to the affected area. Pelagic larvae in marine and freshwater environments enable recruitment from adjacent non-oiled areas. Second, a long life span means a long recovery time. This factor leads to many species of birds and mammals having longer recovery times than fish when the recovery depends on population growth by local reproduction rather than by immigration for other areas. Recovery status and recovery time also depend on how recovery is defined. This observation has been made in reviews of oil spills since the early 1980s through the present day. Other literature reviews all identify problems with defining recovery as a return to historical conditions. These other reviews recommend that defining recovery as a return to a functioning state that provides valuable ecological goods and services is more appropriate. There are cases where recovery may benefit from active human intervention beyond routine clean-up operations. The most protracted recovery appears to be in terrestrial environments where oil reaches groundwater, where cold temperatures decrease biodegradation rates in soil, or when short thaw seasons limit vegetation growth. The state of the art in bioremediation and phytoremediation continues to improve and would appear to be advantageous for spills where natural environmental processes for oil break down are inhibited. Recovery also occurs in the human environments associated with marine, freshwater, and terrestrial environments. Although recovery in the human environments depends on the completion of clean-up activities and recovery of harvested resources such as fish, shellfish, and wildlife, human environments have additional dynamics that shape their recovery. Immediate effects on the human environment derive from the clean-up activities, as well as safety closures and bans for harvesting and recreational areas. Recovery from these immediate effects can take from months to a few years. Examination of the course of recovery following the Exxon Valdez and the Selendang Ayu spills reveals several factors in recovery of human environments (Section 5.2). In the Exxon Valdez spill, protracted litigation and delayed compensation prevented recovery of the human environment for two decades. While the American experience regarding the effects of litigation is instructive, caution must be exercised in extrapolating that experience to a Canadian spill event, which would be subject to different legal processes, rights and entitlements. As explained in the Northern Gateway Application (and Information Responses), ’s Marine Liability Act regime is intended to avoid disputes regarding liability and to effectuate efficient settlement of claims through mandatory insurance and industry-funded pollution compensation fund programs.

July 2012 Page v

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Executive Summary

In contrast, following the Selendang Ayu spill, the settlement process resolved most claims within 5 years and without protracted litigation. Also, in the same spill, a modern Incident Command structure that took human factors into account, and modern spill response capabilities implemented by a dedicated response team were credited with earning the trust of the local residents and reducing effects of clean-up activities on the human environment to less than 2 years.

Conclusions Recovery after oil spills does occur. A review of the scientific literature does not support statements made in Intervener Evidence and oral testimony to the effect that the biophysical and human environments do not recover from oil spills. Recovery is common. The literature is clear that recovery from oil spills is not rare but rather occurs commonly. Recovery from marine spills has received greater study than that from freshwater and terrestrial spills. This review examined 50 spills and 174 Valued Ecosystem components (VECs) from the biophysical environments of cold temperate and sub-arctic regions. Study authors concluded that the VECs achieved a recovered state or recovering state by the study’s end in 86%, 70%, and 50% of the cases for marine, freshwater, and terrestrial biophysical environments, respectively. Another study found 60% of the indicator variable recovered by the study’s end after oil spills in several regions of the world. Where VECs had not recovered, this often reflected the study duration, or complex interactions with other natural or man-caused factors (e.g., killer whale, herring). Recovery takes time. The average time to recovered status for the biophysical VECs examined here was 2.3 and 5.8 years for freshwater and marine environments, respectively. The recovery times for terrestrial VECs are more complicated. Terrestrial VECs that recover appear to do so in about 2 years, but some terrestrial VECs examined here were still recovering after 20 years, or showed no recovery, often reflecting ongoing contamination (e.g., groundwater) and/or lack of appropriate clean-up. Several other reviews have indicated that recovery of marine environments from oil spills takes 2 to 10 years. For human environments, recovery appears to take from 2 to 5 years unless there is protracted litigation. Engagement of communities in determining spill response priorities and developing community mitigation plans can greatly aid in reducing effects and speeding recovery. The time to recovery depends on the environment, VEC, and other factors. Exposed rocky environments recover within a few years whereas sheltered, soft sediment environments such as marshes, can take two decades to recover. Similarly, fast moving freshwater systems tend to recover more quickly than slow flowing freshwater systems. VECs with short life spans can recover within days to a few years but those such as some birds and mammals can take longer. Several factors impede recovery. First, if oil persists, recovery can be slowed. Persistence depends on the characteristics of the oil and the environment. Clean-up is undertaken to remove and lessen the persistence of hydrocarbons. Second, inappropriate clean-up techniques, such as the use of heavy equipment in marshes, can substantially increase recovery time for biophysical environments. Fortunately, modern spill response is based on lessons learned and seeks the most

Page vi July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Executive Summary

appropriate response given the type of environment oiled and other conditions. Third, protracted litigation and delayed compensation can delay recovery of human environments substantially. In a recent U.S. spill, settlement processes without protracted litigation led to resolution of claims within 5 years and with much less social disruption. In some cases, recovery needs active human aid. In terrestrial environments where cold temperature decrease biodegradation rates in soil and short thaw seasons limit vegetation growth, active remediation and restoration may be needed to enhance recovery. Fortunately, the state of the art in bioremediation and phytoremediation continues to improve and would appear to be advantageous for spills where natural environmental processes for oil break down are inhibited. In summary, we quote from Jones and Schmitz (2009) who reviewed recovery from a variety of disturbances: “Our evidence does not support gloomy predictions.”

July 2012 Page vii

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Table of Contents

Table of Contents

1 Introduction ...... 1-1 1.1 Purpose ...... 1-2 1.2 Scope ...... 1-2 2 Behavior of Spilled Oil in the Environment ...... 2-1 3 Background on Recovery ...... 3-1 3.1 Defining Recovery in General ...... 3-1 3.2 Measuring Recovery ...... 3-3 3.3 Processes and Factors in Recovery ...... 3-3 3.4 Human Interventions that Accelerate Recovery of the Biophysical Environment ...... 3-5 3.5 Defining Recovery for Human Environments ...... 3-5 3.6 Summary for Recovery Background ...... 3-8 4 Approach...... 4-1 5 Marine and Estuarine Environments ...... 5-1 5.1 Biophysical Environment ...... 5-1 5.1.1 Marine Water Quality ...... 5-1 5.1.1.1 Exxon Valdez ...... 5-1 5.1.1.2 Braer ...... 5-3 5.1.1.3 Summary for Marine Water Quality ...... 5-4 5.1.2 Shorelines and Sediment ...... 5-4 5.1.2.1 Rocky Shores ...... 5-6 5.1.2.2 Sediment Shores ...... 5-12 5.1.2.3 Saltmarshes ...... 5-14 5.1.3 Plankton ...... 5-17 5.1.4 Benthic Biota...... 5-19 5.1.5 Marine Fish ...... 5-22 5.1.5.1 Herring...... 5-23 5.1.5.2 Anadromous Fish (e.g., salmon, steelhead) ...... 5-31 5.1.6 Marine Birds...... 5-36 5.1.6.1 Pigeon Guillemots ...... 5-38 5.1.6.2 Harlequin Ducks ...... 5-40 5.1.6.3 Other Marine Bird Species and the EVOS ...... 5-41 5.1.7 Marine Mammals ...... 5-42 5.1.7.1 Cetaceans ...... 5-42 5.1.7.2 Pinnipeds ...... 5-49 5.1.7.3 Harbour Seals ...... 5-49 5.1.7.4 Steller Sea Lion ...... 5-51 5.1.7.5 Sea Otters ...... 5-53 5.2 Human Environments Associated with the Marine Environment ...... 5-57 5.2.1 Commercial Fishing ...... 5-60

July 2012 Page ix

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Table of Contents

5.2.2 Traditional Use...... 5-62 5.2.2.1 Recovery of Subsistence Harvesting after the EVOS ...... 5-64 5.2.2.2 Selendang Ayu Harvest Effects ...... 5-68 5.2.2.3 Selendang Ayu Summary of Effects on Traditional Use ...... 5-69 5.2.2.4 Selendang Ayu Subsistence Food Safety and Recovery ...... 5-69 5.2.3 Effects on Food Quality and Safety ...... 5-70 5.2.4 Traditional and Cultural Activities ...... 5-71 5.2.4.1 Defining Traditional and Cultural ...... 5-71 5.2.4.2 Nutrition Transition ...... 5-72 5.2.4.3 Effects and Recovery on Traditional and Cultural Activities Following the EVOS ...... 5-73 5.2.4.4 Summary on Recovery of Traditional and Cultural Activities ..... 5-74 5.2.4.5 Recreation and Tourism ...... 5-75 5.2.5 Passive Use Values ...... 5-78 5.2.6 Archaeological Resources ...... 5-79 5.2.7 Effects on Industrial or Other Land and Resource Uses ...... 5-80 5.2.8 Social Effects ...... 5-80 5.2.9 Effects of Clean-up on Human Environments ...... 5-81 5.2.10 Effects of Compensation and Litigation ...... 5-85 5.2.10.1 Compensation for Direct Damages ...... 5-85 5.2.10.2 Criminal Liability ...... 5-86 5.2.10.3 Civil Liability ...... 5-87 5.2.10.4 Punitive Damages ...... 5-88 5.2.10.5 Summary on Compensation and Litigation ...... 5-89 6 Terrestrial and Freshwater Environments ...... 6-1 6.1 Biophysical Environment ...... 6-1 6.1.1 Soil Quality and Terrestrial Vegetation ...... 6-1 6.1.1.1 Case Studies for Soil and Vegetation ...... 6-2 6.1.1.2 Enhancing Recovery of Soil and Vegetation ...... 6-5 6.1.1.3 Summary on Recovery of Soils and Vegetation from Spills ...... 6-6 6.1.2 Groundwater Quality ...... 6-7 6.1.2.1 Enbridge Line 6b , ...... 6-8 6.1.2.2 Bemidji, Minnesota ...... 6-9 6.1.2.3 Summary on Groundwater Quality ...... 6-10 6.1.3 Freshwater Benthic Organisms ...... 6-10 6.1.3.1 Pine River, British Columbia ...... 6-12 6.1.3.2 Enbridge line 6b Kalamazoo River ...... 6-12 6.1.3.3 Cayuga Inlet, ...... 6-14 6.1.3.4 Asher Creek, Missouri ...... 6-15 6.1.3.5 East Walker River, ...... 6-16 6.1.3.6 Tennessee Pond Trials ...... 6-19 6.1.3.7 Wolf Lodge Creek, Idaho...... 6-20 6.1.3.8 Summary for Freshwater Invertebrates ...... 6-21 6.1.4 Fish and Fish Habitat ...... 6-21 6.1.4.1 Pine River, British Columbia ...... 6-22

Page x July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Table of Contents

6.1.4.2 Enbridge Line 6b, Kalamazoo River, Michigan ...... 6-23 6.1.4.3 East Walker River, California ...... 6-24 6.1.4.4 Reedy River, South Carolina ...... 6-26 6.1.4.5 St. Lawrence River, Quebec ...... 6-27 6.1.4.6 Summary for Fish and Fish Habitats ...... 6-28 6.1.5 Reptiles ...... 6-28 6.1.5.1 Enbridge Line 6b, Kalamazoo River, Michigan ...... 6-28 6.1.5.2 Summary for Reptiles ...... 6-30 6.1.6 Freshwater and Terrestrial Birds ...... 6-30 6.1.6.1 Case Studies for Recovery of Freshwater and Terrestrial Birds ...... 6-31 6.1.6.2 Summary for Freshwater and Terrestrial Birds ...... 6-34 6.1.7 Terrestrial Wildlife and Wildlife Habitat ...... 6-34 6.1.7.1 Grizzly Bears ...... 6-34 6.1.7.2 Mule Deer ...... 6-37 6.1.7.3 Summary for Terrestrial Wildlife ...... 6-38 6.2 Human Environments Associated with Freshwater and Terrestrial Environments ...... 6-38 6.2.1 Human Safety ...... 6-40 6.2.2 Drinking Water and Water Use ...... 6-41 6.2.3 Food Quality ...... 6-42 6.2.4 Commercial Fisheries ...... 6-42 6.2.5 Recreational Activities ...... 6-43 6.2.6 Traditional Use and Cultural Activities ...... 6-44 6.2.7 Industrial or Other Land and Resource Uses ...... 6-45 6.2.8 Effects of Clean-up ...... 6-45 6.2.9 Compensation and Litigation ...... 6-46 6.2.9.1 Summary for the Human Environments Associated Terrestrial and Freshwater Environments ...... 6-47 7 Discussion ...... 7-1 7.1 Recovery Occurrence in Biophysical Environments ...... 7-1 7.2 Recovery Time in Biophysical Environments ...... 7-3 7.3 Factors Influencing Recovery of Biophysical Environments ...... 7-4 7.4 Enhancing Recovery of Biophysical Environments ...... 7-5 7.5 Recovery in Humans Environments ...... 7-5 8 Conclusions ...... 8-1 9 References ...... 9-1 Appendix A Review of Recovery of the Biophysical Environment from Oil Spills ...... A-1 Appendix B List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills ...... B-1 Appendix C Recovery of the Human Environment from Oil Spills: Alaskan Case Studies ...... C-1

July 2012 Page xi

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project List of Tables

List of Tables

Table 3.1 Possible Effects on the Human Environment Associated With Oil Spills ...... 3-7 Table 5.1 Summary table for Exxon Valdez oil spill ...... 5-2 Table 5.2 Summary Table for the Braer Oil Spill ...... 5-3 Table 5.3 Summary Table for the ...... 5-7 Table 5.4 Summary Table for the Arrow Oil Spill ...... 5-8 Table 5.5 Summary Table for the Tsesis Oil Spill ...... 5-9 Table 5.6 Summary Table for the Sea Empress Oil Spill ...... 5-11 Table 5.7 Summary Table for the Oil Spill ...... 5-15 Table 5.8 Summary Table for the Oil Spill...... 5-17 Table 5.9 Summary Table for the Nestucca Oil Spill ...... 5-25 Table 5.10 Summary Table for the Selendang Ayu Oil Spill ...... 5-68 Table 6.1 Summary Table for the Caribou-Poker Oil Spill ...... 6-2 Table 6.2 Summary Table for the Enbridge Line 6b Oil Spill ...... 6-8 Table 6.3 The Feeding Roles of Invertebrate Consumers in Running Waters ...... 6-11 Table 6.4 Summary Table for the Pine River Oil Spill ...... 6-12 Table 6.5 Summary Table for East Walker River Oil Spill ...... 6-17 Table 6.6 Average Fish Density Pre- and Post-the December 30, 2000 Spill ...... 6-25 Table 6.7 Turtles Recovered Following the Enbridge Line 6B Kalamazoo River Spill ...... 6-29 Table 6.8 Summary Table for the Lake Wabamun Oil Spill ...... 6-33 Table 6.9 Summary Table for the Plains Rainbow Pipeline Oil Spill ...... 6-33 List of Figures

Figure 5.1 The Complex Life History of PWS Herring ...... 5-24 Figure 5.2 Spawning Biomass of Pacific Herring in Bay ...... 5-26 Figure 5.3 The Proportion of Total Eggs that were Developed, Hatched, Hatched with Viable Larvae, and Hatched with Larvae having Pericardial Edema ...... 5-27 Figure 5.4 The Settled Volume of Zooplankton in Prince William Sound ...... 5-29 Figure 5.5 Gain in Weight of Adult PWS Herring from Fall to Spring ...... 5-29 Figure 5.6 Spawning Biomass of PWS Herring and Number of Juvenile Pink Salmon Released by Hatcheries into PWS ...... 5-30 Figure 5.7 Adult Pink Salmon Returns to PWS from 1960–2007 ...... 5-35 Figure 5.8 Marine Oil Spill Statistics 1970 to 2011 ...... 5-58 Figure 5.9 Major Marine Oil Spill Events 1970 to 2011 ...... 5-58 Figure 5.10 Location of Major Marine Oil Spill Events ...... 5-59

Page xii July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project List of Tables

Figure 5.11 Annual Numbers of Salmon Harvested by Commercial Fishermen in Prince William Sound from 1976 to 2005 ...... 5-61 Figure 5.12 Estimated Fishing Effort for Prince William Region from 1977 to 2004 ...... 5-77 Figure 6.1 Benthic Macroinvertebrate Taxa Richness (a) and Community Quality Index (b) by Approximate River Mile in Talmadge Creek ...... 6-13 Figure 6.2 Benthic Macroinvertebrate Taxa Richness (a) and Community Quality Index (b) by Approximate River Mile in the Kalamazoo River ... 6-14 Figure 6.3 Illustration of Ephemeroptera, Plecoptera, and Trichoptera (EPT) Taxa by River Mile. Figure a) Represents Cumulative Number of EPT Taxa and b) Represents the Percentage of Sensitive EPT Taxa ...... 6-18 Figure 6.4 Mean Macroinvertebrate Abundance by Location and Time Period ..... 6-19 Figure 6.5 Total PAH (TPAH) Concentrations in the Water Column (a) and in River Sediments (b) by Distance from the Spill and Sample Date ...... 6-24 Figure 7.1 The Number of VECs in Different Recovery Status, by Environment ..... 7-2 Figure 7.2 The Number of VECs in Different Recovery Status, by Study Duration (years) ...... 7-2

July 2012 Page xiii

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 1: Introduction

1 Introduction In written and oral evidence1 before the Northern Gateway Joint Review Panel, it has been commonly stated that spills from the proposed pipelines and tankers calling on the proposed Kitimat Terminal are inevitable. It has also been asserted that the effects of an oil spill – particularly in a marine environment – would be permanent. An example of the arguments advanced to the effect that ecosystems and their Valued Ecosystem Components (VECs) do not recover from spills can be found in submissions filed on behalf of the United Fishermen and Allied Workers Union ("UFAWA"). In it, Dr. Kennedy states, at p. 18: “As indicated previously, Kingston (2002) reviewed the long-term environmental effects of and recovery process for a number of oil spills, including the Exxon Valdez Oil Spill (EVOS), and concluded that in most cases, environmental recovery is complete within 2 to 10 years. This conclusion is controversial and with recent evidence from EVOS (Harwell and Gentile (2006) suggesting that recovery is slower and in some cases, biological communities may never return to pre-spill conditions. Biological recovery times (if at all possible) will be very different for an exposed rocky shore, or for a sheltered shore, or estuarine habitat. Again, biologically sensitive areas may never recover from impacts such as these (Landis 2007).” Aside from expert evidence from individuals such as Dr. Kennedy, many laypersons have stated similar opinions concerning the biophysical environment, either directly to the Joint Review Panel, or in written comments.

1 1. Enbridge Northern Gateway Project, Hearing Order OH-4-2011, Oral Presentation by Haisla Nation, Chief Rod Bolton (Kitamaat Village, British Columbia, Transcripts Volume 8: January 10, 2012) at paras. 3924 - 3925. 2. Enbridge Northern Gateway Project, Hearing Order OH-4-2011, Oral Presentation by Kitsumkalum Indian Band, Chief Don Roberts (Kitamaat Village, British Columbia, Transcripts Volume 10: January 12, 2012) at para. 5213. 3. Enbridge Northern Gateway Project, Hearing Order OH-4-2011, Oral Presentation by Office of the Wet’suwet’en, LAKSILYU clan, Ms. Samantha Vincent (Smithers, British Columbia, Transcripts Volume 11: January 16, 2012) at paras. 5636 - 5638. 4. Enbridge Northern Gateway Project, Hearing Order OH-4-2011, Oral Presentation by Metlakatl Nation, Chief Clarence Nelson (Prince Rupert, British Columbia, Transcripts Volume 20: February 17, 2012) at paras. 11241 - 11242. 5. Enbridge Northern Gateway Project, Hearing Order OH-4-2011, Oral Presentation by Gitxaala Nation, Mr. Warren Nelson (Kitkatla, British Columbia, Transcripts Volume 27: March 13, 2012) at paras. 17363 - 17365. 6. Enbridge Northern Gateway Project, Hearing Order OH-4-2011, Oral Presentation by Saulteau First Nation, Ms. Amy Ann Gauthier (Grande Prairie, Alberta, Transcripts Volume 34: March 28, 2012) at paras. 24838 - 24841. 7. Enbridge Northern Gateway Project, Hearing Order OH-4-2011, Oral Presentation by Heiltsuk First Nation, Chief Gary Housty (Bella Bella, British Columbia, Transcripts Volume 37: April 3, 2012) at paras. 27268 - 27273. 8. Enbridge Northern Gateway Project, Hearing Order OH-4-2011, Oral Presentation by Heiltsuk First Nation, Chief Marilyn Slett (Bella Bella, British Columbia, Transcripts Volume 39: April 5, 2012) at paras. 29156, 29160, and 29212. 9. Enbridge Northern Gateway Project, Hearing Order OH-4-2011, Oral Presentation by Kitasoo/Xai’xais First Nation, Mr. Ernest Mason III (Klemtu, British Columbia, Transcripts Volume 41: April 12, 2012) at para. 30076.

July 2012 Page 1-1

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 1: Introduction

Additional evidence addressed in this document concerns recovery of the human environments. In written and oral evidence before the Joint Review Panel (JRP) for the Enbridge Northern Gateway Pipelines Project (NGP), some interveners have focused on the Exxon Valdez oil spill (EVOS) as their primary “model of choice” for evaluating potential direct, indirect and cumulative effects on the human environment, including effects on Aboriginal people and coastal communities (including non-Aboriginal people); specifically: • Heiltsuk Tribal Council Submission of the Potential Effects of an Oil Spill on the Heiltsuk (A38059) • Gitga’at First Nations Submission Environmental Risk and Effect • Gitga’at First Nations Submission A Social Impact Assessment of the Enbridge Northern Gateway Pipeline Project in Regard to the Gitga’at First Nation • Being Gitga’at: A Baseline Report • United Fishermen and Allied Workers Union Potential Impacts of the Enbridge Northern Gateway Pipeline Project on Members of the United Fishermen and Allied Workers’ Union (Liesel Ashley Ritchie and Duane A Gill) The interveners noted above claim that there will inevitably be a catastrophic marine oil spill, and that this spill will result in permanent social, cultural and ecological damage to coastal First Nations and coastal communities and their way of life.

1.1 Purpose The purpose of this evidence is to reply to the assertion that ecosystems do not recover from spill events. Using a case study approach, scientific literature regarding past spill events is examined to document recovery of ecosystems from events that have occurred in environments similar to that of the study area (e.g. Cold Temperate or Subarctic, rather than Subtropical or Tropical). It will be seen that although oil spills have adverse effects on biophysical and human environments, the scientific literature is clear that ecosystems and their components recover. The literature is also clear that although recovery can occur in some circumstances without clean-up, appropriate clean-up does enable and accelerate recovery. Accordingly, this review also discusses measures such as bioremediation and phytomediation advance recovery.

1.2 Scope The geographic focus of the recovery review is on the north central coast of British Columbia for marine aspects and on the northern area of British Columbia and Alberta for the pipeline aspects. Because of the similarities between the marine areas of the north central coast of British Columbia and the southern part of Alaska, as well as the large extent and duration of studies, specific attention is given to the Exxon Valdez Oil Spill (EVOS). Similarly, the Pine River oil

Page 1-2 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 1: Introduction spill is referenced due to the proximity of this spill to the proposed pipeline right-of-way for Northern Gateway. The term “recovery” as used in this review refers to the recovery of the natural and human environments from an oil spill, not the recovery of oil during the oil spill clean-up response. Because there are many scientific papers and books that review oil spills (Teal and Howarth 1994; Clark 1982; Wells et al. 1995; Rice 1996; API 1999; National Research Council 1995, 2003; Kingston 2002), the review provides only brief descriptions of the effects of the specific spills considered and provides references for further spill details. The main text emphasizes the course of recovery. While recognizing that clean-up response is commonly but not always the first step to environmental recovery, the various case studies considered in this review typically focus on the course of events after the clean-up phase has ended and examines the processes and outcomes of the biophysical and human environments that lead to recovery. Prior to presenting specific information on case studies and recovery of biophysical and human environments, the review first describes the general behaviour of oil in the environment. Definitions of recovery are also discussed.

July 2012 Page 1-3

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 2: Behavior of Spilled Oil in the Environment

2 Behavior of Spilled Oil in the Environment This section provides a synopsis of the behavior of oil spilled into the environment. Substantially greater detail appears in reviews by the National Research Council and others (Whittle et al. 1982, NRC 1985, 1999, 2003). The general framework for understanding oil behavior derives primarily from marine spills of crude oil and some refined products. These spills are the most studied, and the synopsis below reflects that fact. The section concludes with some observations concerning spilled oil behavior in terrestrial and freshwater environments. At the outset, it is emphasized that crude oil is a mixture of natural chemicals. Although the modifications that are made to crude oil in refining to produce particular products can significantly alter its fate and behavior in the environment, the vast majority of the petroleum- based aromatic and aliphatic compounds that are found in these products are subject to the same degradation processes as the original parent compounds. As soon as oil is spilled into the terrestrial, freshwater, and marine environments numerous chemical, physical and biological processes immediately begin to break down, biodegrade and otherwise assimilate the spilled oil (Whittle et al. 1982, NRC 1985, 1999, 2003). This natural degradation of oil sets the conditions under which the recovery of the biophysical and human environments from oil spills occurs. Ultimately, spilled oil is broken down into carbon dioxide and water by sunlight (photolysis) and microbes (biodegradation). However, the time required depends on the nature and amount of the petroleum product spilled and the characteristics of the receiving environment. Under calm conditions, oil spilled onto water surfaces has a tendency to spread by gravity into a thin layer (about 0.1 mm), covering a relatively large area. However, in the environment, this tendency is modified by the actions of wind, tides, and currents, resulting in an uneven distribution of oil slicks of various thicknesses, along with other changes. These slicks can be transported by wind and current to strand on shorelines. As the slicks weather, the oil viscosity increases and the tendency to spread decreases. Under the influence of winds, many oils form stable water-in-oil emulsions called mousse. Mousse is the sticky, dark coagulum that accumulates along shorelines. Because of the amount of water incorporated in the mousse, mousse formation increases the amount of contaminated material that has to be dealt with. Strong winds also force oil into solution and increase the rate of evaporation. Different weathering processes dominate at different times after spillage. Evaporation acts within hours to move most of the volatile and toxic oil components into the atmosphere, where sunlight oxidizes the compounds. Depending on the oil type and the content of volatile compounds, evaporation can remove up to 50% from the water surface. Dispersion occurs when wind waves or other turbulence drives oil into the water column where it is present as droplets. When the waves or turbulence abates, these oil droplets can coalesce into larger droplets that float back to the water surface. The terms “dispersed,” “sunken,” and “submerged” oil need clarification (NRC 1999). Naturally dispersed oil is oil that is driven into the water column by some form of turbulence, such as that resulting from waves in marine and lake environments or high, turbulent flows in rivers. This

July 2012 Page 2-1

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 2: Behavior of Spilled Oil in the Environment

dispersed oil is present as droplets that can sometimes coalesce and resurface when the turbulence decreases. Oil can also sink by gravity when its density is greater than that of the receiving water. Oil density can increase when weathering removes the lighter volatile compounds or more commonly when oil contacts sand or other particulate fines and increases in density through the sorption of these heavier particles. Because distinguishing the mechanisms that have operated in a given spill can be difficult, the NRC (1999) recommends using the term “non-floating oil.” This term is used in this document. Oil spilled into the terrestrial environment spreads by gravity but the spreading rate and pathway is heavily influenced by the terrain, soil types, and vegetation. The tendency of oil to seep into soil and migrate into groundwater is affected by soil characteristics such as grain size and by oil characteristics such as viscosity and weathering state. Oil spreading over land tends to pool in low areas in the terrain. Oil spilled on land can reach freshwater. Oil spilled into freshwater systems, such as streams, rivers, ponds, and lakes, is subject to the same processes of evaporation, dispersion, dilution, dissolution, and transport as described above for oil in marine systems. The turbulence associated with swift currents in streams and rivers can physically disperse oil into the water column. As in marine systems, the dispersed oil droplets can coalesce when the turbulence decreases and the larger droplets float to water surface. Because of the shallow depths of rivers compared to marine waters, there is greater opportunity for the dispersed oil to reach bottom substrates and sorb particulate matter. Freshwater is less dense than marine waters and the sorption of particulates can increase the density of heavy oils to the point that it sinks by gravity. Ultimately, most of the oil spilled into terrestrial, freshwater and marine environments is broken down by natural agents. With no clean-up or other treatment, about 75% of crude oil spilled into the marine environment will be broken into carbon dioxide and water (Whittle et al. 1982). Biodegradation rates depend on the oil type and characteristics of the receiving environment, such as temperature and prevailing microbial populations (NRC 2003).

Page 2-2 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 3: Background on Recovery

3 Background on Recovery Biophysical and human environments do recover from oil spills and other disturbances but characteristics of the environment (e.g., petroleum degrading microbe populations, nutrients, temperature, salinity) and the oil (e.g., heavy vs. light) both influence the time to recovery. Perceptions and conclusions often differ because of the way people have defined and assessed recovery. This section discusses how recovery has been defined and assessed in a number of oil spills. The section also describes the processes involved in recovery and the measures that are often taken to accelerate or enhance recovery. The bulk of the section focuses on the biophysical environment and ecosystems. Recovery in human environments is also discussed. First, it is important to distinguish restoration from recovery. Restoration is the active human intervention that enables or accelerates recovery. Ecological recovery, then, includes the processes that return the ecosystem to a desired functioning state. As discussed elsewhere in this document, the physical collection and removal of oil from the environment can be an important first step in enabling and accelerating natural recovery.

3.1 Defining Recovery in General Definitions of ecological (biophysical) recovery have both regulatory and scientific contexts, and in both, the definitions have changed substantially over the last several decades. The common element in most definitions is a return of the ecosystem or a particular valued ecosystem component (VEC) to some desirable system state following a disturbance. This state can be a previous historical state or a functional state where the ecosystem or VEC is once again producing valuable ecological goods and services. As background to any definition, one must note the dynamic, ever-changing and highly variable character of natural systems. To quote two statements from two noted ecologists (Jackson and Hobbs 2009): “First, environmental and ecological changes are normal; perhaps the most natural feature of the world in which we find ourselves is its continual flux.” “In the long run, no inherent [particular] natural ecosystem or landscape configuration exists for any region.” It is within these continuously changing environments that oil spills occur. Also, it is within this continual change and multiple possible pathways and ecosystem states that recovery from oil spills runs its course. The ways in which recovery has been defined include return to: • natural equilibrium • previous historical conditions • pristine conditions

July 2012 Page 3-1

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 3: Background on Recovery

• some notional baseline state • natural variability from comparison of affected and reference areas • conditions that “might have been but for the spill.” • functional ecosystem state that provides valuable ecological goods and services In their growing recognition of the natural dynamics of ecosystem change, ecologists have moved away from using such recovery end states such as return to equilibrium (Landis 2007; Wu and Loucks 1995; Kapustka and Landis 1998; Jackson and Hobbs 2009). Jackson and Hobbs (2009) argue that return to a previous historical state is problematic because our understanding of past historical system states is inadequate. Also, many previous historical states cannot be achieved because other historical natural and man-made events have fundamentally changed the system. Similarly, the notion of a return to pristine conditions is problematic because pristine conditions are too often an assumption rather than a reality. For example, the Prince William Sound ecosystem, despite prevailing beliefs, was not pristine at the time of the EVOS (Wooley 2002). Extensive ecosystem transformations had already occurred through commercial harvesting of fur, fish, and timber, as well as mining and the introduction of non-indigenous species. Also, just prior to the EVOS, harbour seal and Steller sea lion populations were in decline (Braham et al. 1980; Calkins et al. 1994; Frost et al. 1994) and hatcheries were adding over 500 million juvenile pink salmon annually (Pearson et al. 2011). Return to baseline often involves a comparison of post-spill conditions with “baseline” conditions measured usually for a single VEC species or species guild during some pre-project phase of development. Where such baselines have been based on only a year or two of study, they have proven to be inadequate, because basic system dynamics can require years to fully assess. Defining recovery as a return to the natural variability of the resource was intended to address some of the shortcomings of the comparison-to-baseline approach. This definition assumes that recovery occurs when the resource in the affected area is “tracking” (i.e., changing in parallel with) the changes in the resource in an unaffected reference area. More details on the ways in which this definition can be assessed are given in the subsection below on measuring recovery. Recently, recovery has been defined as a return to the conditions that would have prevailed had the oil spill not occurred. This definition recognizes the need to account for natural variability and for the influence of natural and man-made factors other than the spill. Jackson and Hobbs (2009) also argue that the goal of recovery and of active restoration of damaged environments should be to recover or restore the ecosystem to a functional state that provides valuable ecological goods and services. This perspective is also emerging in the regulatory context in North America.

Page 3-2 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 3: Background on Recovery

3.2 Measuring Recovery Two reviews have compared study designs and ways to measure recovery (Wiens and Parker 1995; Parker and Wiens 2005), including: • comparison to baseline • single year comparison of affected versus reference area • multiyear comparisons of affected versus reference areas • Before-After-Control-Impact (BACI) studies • weight of evidence Parker and Wiens (2005) concluded that there is no one “best” way to assess recovery. They urge that future oil spill investigators examine how the ecosystem to be studied functions, determine which set of assumptions prevail for that ecosystem and then select an appropriate study approach. In recent years, the BACI approach often has been favoured for assessing oil spill recovery (Parker and Wiens 2005). Landis (2007) compares the different conclusions about recovery from the EVOS in two prominent papers (Peterson et al. 2003, Harwell and Gentile 2006) and seeks to understand why the conclusions are so far apart. Landis attributes part of the differences between the two papers to “the infusion of social values or policy goals into each.” However, for Landis (2007), the more important source of the differences between the two papers is how they differ in their approaches to defining recovery and the role of indirect effects and effects cascades. Indeed, Peterson et al. (2003) do not offer an explicit definition of recovery and appear to imply that recovery is the return to some undefined previous state. In contrast, Harwell and Gentile (2006) do offer a recovery definition – return to an acceptable range of variability in the resource of interest – but one that includes a policy statement (“acceptable”) and one for which the outcome rests on a judgment of acceptability rather than on a scientific operation such as a statistical comparison of the attributes of two areas or populations. Landis (2007) recommends that researchers provide or come to consensus on clear and operational definitions and goal statements in scientific assessments of effects and recovery.

3.3 Processes and Factors in Recovery Any discussion of the processes in recovery from oil spills must first recognize the critical role that clean-up plays in enabling and accelerating recovery. One lesson learned from the comprehensive review by Borja et al. (2010) in over 50 case studies of the recovery of estuarine and coastal ecosystems from environmental degradation is that the stressor or disturbance agent needs to be reduced to some extent before recovery can progress. A corollary from the same study is that both human interventions such as oil spill clean-up and natural processes such as biodegradation must reduce the contamination of the system but does not need to fully eliminate all contamination before recovery can progress.

July 2012 Page 3-3

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 3: Background on Recovery

The natural processes that advance ecological recovery include: • Natural transport – wave action, currents, and other processes physically move the oil from one environmental compartment to another • Natural transformation – photolysis and biodegradation break down the oil • Natural population growth – the population rebuilds through natural recruitment and growth • Immigration, colonization, and recolonization – organisms move into the affected area and re-establish their populations • Succession – naturally occurring waves of immigration and colonization lead to successive waves of species and communities following each other The review of Borja et al. (2010) provides insights into the factors that influence the course of recovery. The energetics of the system govern the rate at which contaminants are transported or transformed. Beaches with high-energy wave action tend to quickly clean themselves, while sheltered wetlands and embayments tend to retain oil contaminants longer. The season when the spill occurred can influence the nature of the effect and the time course of recovery. When a spill occurs during the season for colonization of invertebrates, such as mussels and clams, recovery may be delayed for another cycle. The turnover or lifespan of the wildlife receptor of interest can also influence the time course. Phytoplankton and zooplankton that have very short lifespans and high turnover rates (some on the order of days) tend to show little effect and rapid recovery. Long-lived species, such as birds and mammals, may have long recovery times. The extent to which the species has widespread dispersal of eggs or larvae can influence the spatial extent of recovery. The fecundity or reproductive potential of the resource can also influence the rate of population growth. The Borja et al. (2010) review of 53 degraded environments compares the time course of recoveries for different types of stressors. Oils spills have moderate recovery rates. About a half of the stressor categories show recovery in less than five years, some in less than two years. Oil spills showed recovery in two to ten years. More extensive or intensive disturbances, such as disposal of sewage sludge and mine tailings, land reclamation, and long-term wastewater discharge, have shown recovery over ten to twenty years. A widespread observation among investigators studying the effects oil spills and other disturbances is that the longer the time taken in recovery, the greater the risk that some natural event or other man-made factor will influence the course of recovery or active restoration (Parker and Wiens 2005; Jackson and Hobbs 2009; Borja 2010).

Page 3-4 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 3: Background on Recovery

3.4 Human Interventions that Accelerate Recovery of the Biophysical Environment Activities that can enable or accelerate recovery include: • cleanup • remediation and bioremediation • harvest moratoria • restocking through transplants from other unaffected areas or from hatcheries or nurseries • active restoration of habitat The aim of cleanup is to reduce the level of contamination to the point where natural factors will complete the process of recovery or to where remediation and bioremediation can be used to further reduce contamination and speed recovery. For commercial species or other harvested species, harvest moratoria and restocking can accelerate recovery by reducing or counter- balancing mortality from other sources and rebuilding populations more quickly. Where habitat is limiting growth and survival, active restoration can improve the quality or quantity of habitat to increase population growth or recolonization.

3.5 Defining Recovery for Human Environments A review of literature related to the effects of terrestrial, freshwater and marine oil spills shows that there have been few attempts to define criteria that can be used to determine when human environments have recovered from the effects of a spill. In general terms, each spill event is unique and recovery efforts for the human environment appear to focus on cleaning up the immediate effects of the spill to allow some human activities to resume, restoring ecological functionality and allowing human activities to return to near normal, and providing financial compensation and assistance offset any losses that may have occurred. As the assessment is discussed later in this review, each of these three types of recovery efforts can, in turn, have its own effects on human environments and may have additional implications for recovery. For this analysis, we have described recovery of the human environment from the effects of a spill in terms of four general “phases” that reflect environmental factors as well as regulatory factors: • The spill. These would include the initial effects of the spilled product itself on the ability of people to use the natural resources affected by the spill, such as beach closures, sport and commercial fishing restrictions, shellfish harvesting limitations, etc. In some cases, this includes evacuation of people living nearby and potential health effects associated with direct exposure to the oil or associated air emissions. These effects tend to be very short-term. • Clean-up effects. These activities typically involve implementation of an approved response plan and efforts to restore the natural and human environments to the extent required by regulations and that is physically and financially practical. Human environments can be affected by both activities related to spill assessment and spill response. Of particular interest

July 2012 Page 3-5

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 3: Background on Recovery

is the extent to which local populations participate in response activities and the number of additional response personnel and materials being brought into local communities. Importing response personnel can result in temporary demographic changes within communities and can place additional demands on community infrastructure and services. Response activities typically involve removing as much oil as possible, cleaning of infrastructure (such as boats, docks, etc.), treating the environment to the extent practical, and essentially allowing day-to- day activities to return to near normal. This means that companies can return to withdrawing water for commercial or industrial uses, boats can travel through the area, commercial fishing can resume, etc. While the economic activity associated with response activities can be seen as a temporary replacement for some economic activities that were curtailed as a result of the spill (and be a benefit for the overall community), there may be uneven distributional effects (the perception of winners and losers within the community). Response activities can often be achieved relatively quickly, usually within one year, although clean-up of larger events can take longer, especially if winter conditions prevent a year-round response. • Return to ecological functionality. Over the longer term, the recovery of the human environment will depend on the time it takes to restore ecological functionality in the biophysical environment, which allows individuals and communities to return to the land and resource use harvesting activities much as they were prior to the spill, without having to worry about the effects of residual oil or treatment effects. For example, a spill on agricultural land could be cleaned up the first year, but the return to planting and harvesting may be delayed for another year or two while the land and resource base recovers. Similarly, the return to commercial and traditional fishing may be delayed until fish populations return to pre-spill levels and food testing indicates there are no food safety concerns. This stage of recovery is especially important for marine/aquatic resource based communities. • Restoration. The final stage of recovery involves restoring the well-being of individuals and communities adversely affected by the spill and relates to the payment of financial compensation for damages. Once all accounts have been settled and compensation has been paid, the financial aspects of the human environment are technically considered to have recovered. The time required to fully restore well-being depends on the processes by which compensation is paid. In some cases, compensation arrangements are negotiated and implemented cooperatively and can occur relatively quickly. However, in some cases, the final resolution of compensation may require deliberations by the courts, and this process can be extremely lengthy. In the case of EVOS, it has been argued that 20 years of litigation formed a barrier to gaining a coherent picture of recovery. While some researchers with long term experience in the area noted that communities affected by EVOS had made substantial recovery within five years (Davis 1996), the effects of the spill were exacerbated by the lack of synthesis and coordination of the various agencies conducting socio-economic research, the “gag orders” that were initiated in communities by class action lawyers, and other barriers to objective social science. And, even once well-being may have been technically restored through the payment of compensation, some individuals still may not be entirely satisfied with the outcome.

Page 3-6 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 3: Background on Recovery

The following examples have been provided to show the types of activities and possible effects on the human environment associated with oil spills in different environments.

Table 3.1 Possible Effects on the Human Environment Associated With Oil Spills Functional Compensation and Resource Spill Damages Clean-up Recovery Litigation Fisheries Clean-up of Hiring of vessels Fish populations Damages paid for loss resources vessels, and crew to assist recover and no of income during fish (commercial) replacement of tainting so closure nets/gear commercial fishing can resume Traditional and Health warnings Acceptable Resource Compensation paid to country food and closure of replacement food harvesting can communities, culture harvests fisheries sources provided occur with no camps supported, etc. (personal / concerns about community) resource health or access restrictions Ceremonial Health warnings Acceptable Harvests occur Elder and traditional harvests and closure of substitute sources with no concerns knowledge holder areas to provided from about resource involvement in harvesting adjacent areas or health recovery, culture activities groups camps supported, etc. Drinking water Provision of Resume near Normal operations Damages paid for alternate short normal operations resume with no effects on term water supply concerns about infrastructure and residual oil creation of alternative supply sources Commercial or Plant shut down Temporary fixes Water can once Plant operators receive industrial water allow operations to again be diverted compensation for users resume with no concerns costs/loss of revenues about residual oil for period of closure Recreation Closure of Treatment of Beaches and Adjacent landowners beaches beaches and features become compensated for lack recreation open and of access during the infrastructure available for use spill allows activities to with no concerns re-open about residual oil Tourism Restrictions on Tourism facilities Tourism numbers Businesses receive access to affected used by recover to near compensation for loss areas responders normal of use Private property Immediate Clean-up of Residents return Residents receive evacuation and property compensation for loss temporary housing of use Agricultural Cessation of Clean-up of Agricultural Famers receive land agricultural property allows production returns compensation for loss activities some limited to near normal of income agricultural use

July 2012 Page 3-7

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 3: Background on Recovery

It should be noted that the literature review provided only one example where specific recovery objectives were established for the human environment. A major portion of the natural resource damages settlement paid by Exxon in the case of EVOS was placed in a trust fund that was to be used by the EVOS Trustees Council (EVOSTC) to fund restoration projects. The EVOSTC developed various recovery objectives as a means of demonstrating how these funds were being used. Specific recovery objectives were developed for commercial fishing, recreation and tourism, subsistence activities, passive use values, and archaeological resources. These objectives and the current status of recovery for these human services are discussed in Section 5.2. In general, recovery was defined in terms of the return to pre-spill conditions, but this approach has proven problematic because of a lack of knowledge of pre-spill conditions in the affected area and the realization that activities tied to resource harvesting are affected by natural resource cycles as well as external factors that affect the demand and supply of resource products.

3.6 Summary for Recovery Background The common element in most definitions of recovery is the concept that the ecosystem, the VECs, or the human environment must return to some more desirable state. The most recent definitions take into account how the ecosystem is changing because of natural and other man- made factors and, therefore, include concepts similar to return to the conditions that might-have- been, except for the spill. In the analysis of the case studies that follow, this review attempts to answer questions about the processes that have been involved in recovery, along with the influence of any obstacles and human interventions. Two concepts dominate this report’s definition of the recovery of the biophysical environment: • return to the state that would have been if not for the spill • return to a functional ecosystem that provides valuable ecological goods and services Similarly, recovery of the human environment occurs when the human system returns to the state that would have been if not for the spill and to a functional state providing valuable cultural, social and economic goods and services.

Page 3-8 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 4: Approach

4 Approach Two basic approaches were employed in preparing this review. First, the scientific literature in peer reviewed journals and government reports was reviewed to identify and acquire information on recovery from oil spills. Second, from the oil spills where recovery was followed, case studies of the recovery of specific species, groups of species, human uses and human values were selected using a set of a priori criteria. Because the review focuses on recovery rather than on effects, the selected case studies need to have been studied sufficiently well to assess whether recovery had occurred. For each oil spill and each Valued Ecosystem Component (VEC; for example fish and fish habitat) that were examined as candidate studies, several questions were asked: • What VEC was affected? • What was the reported effect? • Was recovery reported? • If the resource recovered, what processes were involved? • Were there obstacles or complicating factors that impeded recovery? • Were there attempts to intervene to accelerate recovery? • How long did recovery take? The results of this process are tabulated in Appendices A and B. From this set of candidate oil spills, specific spills and VECs were selected for more detailed treatment as case studies. The criteria for selection were: • Cold Temperate Zone or Subarctic: The spill was in a cold temperate zone or subarctic zone, and not a tropical or subtropical zone • Recovery Information: The spill received sufficient study to follow at least some aspect of recovery. • Similar environment: The spill occurred in an environment similar to that of the Project Area. • Only spill with information: The spill is the only one with information on an aspect of concern. The level of information may fall short of what might be hoped for, but the spill was included because of its overriding interest. Assessment techniques were noted where they were clearly explained. Generally, the appendix gives the state of recovery reported by the study authors. For many spills, the study ended before full recovery was achieved. Such studies were scored as recovering and the study duration since the time of the spill was listed. Partial recovery was listed where the study authors indicated that only partial recovery had occurred.

July 2012 Page 4-1

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

5 Marine and Estuarine Environments

5.1 Biophysical Environment

5.1.1 Marine Water Quality The concentration of oil in the water column under an oil slick is generally low (Kennedy 2002). After oil is spilled, evaporation of the volatile components and dispersion of oil into the water column removes oil from the water surface. Natural wave action and turbulent currents can disperse oil into fine droplets (0.01 to 1 mm in diameter). Oil in the water column has one of three fates. First, hydrocarbons from the dispersed oil and slick dissolve into the receiving water to be diluted and rapidly degraded by microbial action. Second, after the waves or currents abate, the oil droplets coalesce into larger droplets that float back to the water surface. Third, the oil droplets accumulate sand particles suspended in the water column and then sink. The EVOS and the Braer oil spill are well-studied examples where the recovery of water quality was followed closely.

5.1.1.1 Exxon Valdez When the Exxon Valdez oil spill (EVOS) (Table 5.1) occurred in March 1989, it was the largest release of crude oil (11 million ) that had ever occurred in US waters (Wells et al. 1995). Marine species in the affected waters could be exposed to oil in the water column or residual EVOS polycyclic aromatic hydrocarbons (PAHs) in near shore and intertidal water waters (water that flows over the shore during the tidal cycle). Potential routes of exposure included drinking small amounts of seawater with their food or by fouling their fur or feathers with dissolved and dispersed oil or oil sheens. The health risk posed by such exposure depends on the concentration and duration of exposure to Total Polycyclic Aromatic Hydrocarbons (TPAHs) in surface waters. More than 2000 water samples were collected throughout PWS and analyzed for TPAH from shortly after the spill in 1989 to 2005. More than 2,700 indigenous and deployed mussels were also analyzed for PAH in this time period from oiled and unoiled shores to calculate PAH concentrations in water from PAH residues in mussel tissues (Neff and Burns, 1996). These data were used to document long-term trends in TPAH concentrations in the upper water column of spill-path and non-spill-path areas of PWS (Boehm et al. 2007). By early 1990, TPAH concentrations in the water column and in water washing the intertidal zone within the spill path had declined to concentrations similar to those in surface waters outside the spill path in PWS (<0.05 µg/L) and well below levels that might pose a risk to marine organisms (Neff and Stubblefield, 1995).

July 2012 Page 5-1

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Table 5.1 Summary table for Exxon Valdez oil spill Oil Spill Name Exxon Valdez Location Prince William Sound, Alaska Year 1989 Oil Type Crude Specific Name Crude Volume (metric tonnes) 33,000 Platform Tanker Environments Affected Marine Valued Ecological Component (VEC) Studied • Algae • Birds • Fish • Macroinvertebrates • Mammals • Sediment • Shoreline • Water quality Summary: On March 24, 1989 the tanker vessel Exxon Valdez grounded on Bligh Reef and spilled approximately 33,000 metric tonnes of crude oil into Prince William Sound Alaska, USA. Approximately 10,000 km of coastline was affected. The estimated volume of recovery from the initial cleanup efforts was less than 10 percent. Extensive cleanup of affected beaches occurred mainly in 1989 to 1991 with the final cleanup in spring of 1992. Studies of a number of VECs were followed sufficiently long for recovery to be evident.

Short et al. (2007) estimated that the amount of oil residues on the shore was decreasing by about 4%/y after 2001 and hypothesized that most oil loss was by dissolution/ dispersion of oil from shoreline residues into tidal water, particularly during storms. The rate of leaching of PAH from intertidal oil residues depends on 1) the extent to which tidal water (including sediment pore water, sea water overlying the shore at high tide, and fresh water runoff) contacts the oil residues, either during the tidal cycle or during storms that expose them through the displacement of surface rock armour and 2) the extent of weathering of shoreline oil residues. The oil residues washing off the shore are present in the near shore water as dissolved and dispersed hydrocarbons in the water column and as surface sheens. The slow estimated rate of loss of shoreline deposits from oiled shores after 2001 (~ 4%/y), the high rainfall (150-635 cm/y), and the large tidal prism in PWS (>3 m) ensure that any EVOS residues dissolving or dispersing from intertidal sediments into near shore waters through natural processes will be diluted rapidly to background concentrations. TPAH concentrations in near shore waters in 2002 through 2005, based on analysis of water samples and estimation of concentrations in water from residues in indigenous, intertidal mussels, were similar in previously oiled and unoiled areas of PWS (Boehm et al. 2007a). Concentrations ranged from 0.001 to 0.57 µg/L at oiled sites and <0.001 to 0.25 µg/L at unoiled

Page 5-2 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

sites. This result indicates that the PAHs are being mobilized to a very limited extent from intertidal surface and subsurface oil residues and are being diluted rapidly to background levels in the near shore water column. The low PAH concentrations in near shore water are confirmed by low levels of cytochrome P450 mixed function oxygenase (CYP1A) activity in territorial intertidal fish from oiled and unoiled shores (Huggett et al. 2006). Huggett et al. (2006) found that hepatic CYP1A activity in high cockscomb pricklebacks (Anoplarchus purpurescens), a fish with high site fidelity that lives among rocks in the lower intertidal zone in PWS, was low and at identical levels at unoiled sites and sites with residual subsurface oil. Earlier, Jewett et al. (2002) found no correlation between CYP1A activity in two species of near shore fish, the masked greenling (Hexagrammos octogrammus) and crescent gunnel (Pholis laeta), and petroleum hydrocarbon concentrations in sediments at formerly oiled sites 7 to 10 years (1996-1999) after the spill. These observations indicate that the bioavailability of PAH from shoreline oil residues is very low. All measured or estimated TPAH concentrations in near shore waters of PWS between 1990 and 2005 are orders of magnitude below the state of Alaska marine water quality criterion for total aromatic hydrocarbons (including both monocyclic and polycyclic aromatic hydrocarbons) of 10 µg/L (Scannell et al. 2005). Thus, there is a negligible risk that marine species are being exposed to toxicologically meaningful concentrations of EVOS PAH in the water column.

5.1.1.2 Braer In the Braer spill (Table 5.2), extreme weather conditions with high winds and waves dispersed the majority of the spilled oil into the water column (Kinsgton 2002). Over two weeks of high winds (average 10 m/s with maximum of 16 m/s) with long fetch produced high breaking waves (Thorpe 1995). Within a few days, hydrocarbon concentrations peaked at over 1500 ppb. However, recovery from these high water column concentrations occurred in a matter of days. Hydrocarbon concentrations in the water column decreased from the peak concentrations to less than 100 ppb within 10 days and to low ppb levels within 40 days.

Table 5.2 Summary Table for the Braer Oil Spill Oil Spill Name Braer Location Islands, UK Year 1993 Oil Type Crude/No. 6 Specific Name none Volume (metric tonnes) 84,000 Platform Tanker Environments Affected Marine

July 2012 Page 5-3

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Table 5.2 Summary Table for the Braer Oil Spill (cont’d) Valued Ecological Component (VEC) Studied • Fish • Macroinvertebrates • Shoreline Summary: On January 5, 1993 the tanker Braer ran aground at Garth Ness on the southern Shetland Isles and spilled approximately 84,000 metric tonnes of crude/no. 6 fuel oil. The affected areas were concentrated south of Fair Isle and west of Burra Isle totaling approximately 4,080 km2 area. The estimated volume of recovery from the initial cleanup efforts was could not be found in the literature. Studies of fish, macroinvertebrates, and shorelines were followed to address the progress of recovery.

5.1.1.3 Summary for Marine Water Quality Generally, hydrocarbon concentrations in marine waters are approximately a few ppb and decline from the natural processes of dilution and biodegradation. Even in extreme cases, such as the Braer spill where high winds and waves dispersed oil into the water column, return to low levels occurs within days to weeks.

5.1.2 Shorelines and Sediment There is a considerable body of literature on the effects of oil spills and oil spill clean-up on intertidal habitats and species. Shorelines are vulnerable to contamination from surface oil, and are relatively accessible compared to the sub-tidal. Studying effects on intertidal communities can be more readily accomplished because many intertidal organisms live in or attached to the substrate and are easy to observe. The majority of reports and published papers describe the initial effects and many describe the status of recovery after one or two years post-spill. At that point, many studies stop because of lack of funding, or the recovery appears to be complete or well underway, or a lack of pre-spill data limits the value of continuing. A few studies have continued, often with long periods of inactivity, when funding is available and the affected resource has shown signs of long-term injury. This subsection focuses on the longer-term studies, but also includes examples of the many cases where the scale of effect has been small and the recovery rapid. The following review documents are particularly relevant to this section: Baker et al. 1990, AURIS (1994 and 1995) and Baker et al. (1996). Other more general reviews of effects and recovery in cold water marine environments include Sloan (1999), Mosbech (2002) and AMAP (2008). Sell et al. (1995) (based on AURIS 1994) reviewed 64 studies of oil spills on rocky shores and saltmarshes and found that the majority of spill affected communities underwent a rapid natural recovery process without requirement for clean-up. They identified three stages of the typical recovery process: 1) initial colonization (characterized by the conspicuous settlement and growth of macroscopic opportunists), 2) recovery (during which the community progresses toward, but has not yet attained its natural range of dominance, diversity, abundance and zonation, and 3) recovered (where a natural biota has been established and is within the range of dominance, diversity, abundance and zonation expected for that habitat). It is a feature of all coastal oil spills that shoreline oiling is patchy, even after large spills, so that within the stretch of coast that receives the majority of the oil there are normally large areas that are only lightly oiled and

Page 5-4 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments substantial areas that are not physical oiled and will be affected only by elevated concentrations of hydrocarbons in the water. Moore (2006a) reviewed the factors that can result in slow recovery and concluded that there were essentially four: • The persistence of oil - Exposure to wave action greatly reduces the persistence of oil and is arguably the most important element in natural clean-up and a major pathway by which affected ecosystems recover. It is the lack of water movement on very sheltered (usually <20km fetch) shorelines that is the main cause of most of the long-term ecological effects. While oil characteristics and environmental conditions significantly influence the fate of oil that reaches very sheltered habitats, the inherent vulnerability and sensitivity of those habitats is high. The persistence of oil in shoreline habitats that are more exposed to wave action is considerably less, and it is unusual for substantial amounts of toxic or smothering oil to remain for many years. However, this can happen if heavy oiling occurs during a sufficiently long period of calm seas for tar to form intractable residues; mixed sand/shingle shores are particularly vulnerable to asphalt pavement formation. The physical smothering by tar residues can effectively reduce habitat diversity by binding substrata and filling spaces, which can produce the most conspicuous long-term effects on biodiversity and productivity. The chemical toxicity of weathered tar is much reduced by physically locking it up inside a deposited tar ball or coating. Many such deposits (particularly in upper shore rock and shingle habitats) have very little ecological effect, and observations of limpet grazing marks and algae colonizing the surface of old tar residues has been observed at a number of historical spill sites. Young oil deposits are clearly not completely benign and long-term sheening can even occur from old deposits; but it is the physical presence of the oil that has the main effect. However, hydrocarbons trapped in sheltered sediment habitats can be more bioavailable and cause more long-term effects through chemical toxicity, as shown by some bioassay tests and some biological effects studies. Degradation and reduction of the toxicity takes place over time, but in the most sheltered anoxic muds this may take more than 20 years. • Slow growing, long-lived species - Mortality of slow-growing long-lived species is likely to cause at least some longer term effects on population structure, even if there are no other impediments to recovery. Particular concern is for affected species that have a major structural role in the community they live in (by physical size or other strong ecological influence) and those that have an important ecosystem role (e.g., through production of organic matter that may be exported widely). Vulnerable groups of shoreline species with life spans over 10 years include perennial saltmarsh plants, lichens, encrusting coralline algae, some kelps and other brown algae, some crabs and lobsters, some snails (particularly limpets), some clams and burrowing urchins. • Limited potential for recruitment - It is unusual to get wide broad-scale loss of any species, because oil distribution is normally very patchy and much less than the scale of dispersal of most species. However, some species have localised populations that are geographically isolated from sources of outside recruitment or are otherwise characterised by limited

July 2012 Page 5-5

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

recruitment. Factors include physical barriers, distance, edge of distributional range, strong linkages between breeding and feeding sites, limited dispersal mechanisms for spores, larvae, juveniles or adults, or other features of reproduction that limit local recruitment. • Severe clean-up actionconanhe removal of bulk oil that is causing acute effects can also reduce long-term effects. However, physical clean-up actions can cause long-term effects if applied very severely. The spill studies provide examples of long-term effects from severe clean-up in marshes, but the examples from rocky and boulder shores (Torrey Canyon and Exxon Valdez) suggest that the primary-recovery processes can be rapid and that long-term effects consist of abnormally high fluctuations in the affected populations that dampen over time.

5.1.2.1 Rocky Shores The sensitivity and recovery of rocky shore communities from oil spills depends mainly on the degree of exposure to wave energy. Exposed rocky shores are normally one of the habitats that are least vulnerable to oil spills, because the oil is quickly removed by wave action. However, sheltered rocky shores can be more vulnerable and sensitive if they include numerous rock pools and crevices. A number of conclusions about sensitivity and recovery can be drawn from the various studies of oil spill effects on rocky shores in cold temperate and sub-arctic zones. Acute mortality of intertidal limpets is a good indicator of fresh oil contamination (by liquid oil or very high concentrations dispersed in water), but mortality is much reduced if the oil is weathered. Because they are long lived and have an important structuring role in many rocky shore communities, restoration and recovery of such communities often depends on their recolonization and the re- establishment of a natural age structure. Other gastropods appear to be less sensitive to oil than are limpets, but they may still suffer some acute mortality if large amounts of fresh oil or high concentrations of oil are present in the water. Diversity and abundance of small crustacea (e.g., in kelp holdfasts and algal turf habitats) are greatly affected by hydrocarbon concentrations in water (and presumably by liquid oil). They typically have multiple generations per year, so recovery is likely to be rapid if there are no barriers to their recolonization. Mortality of barnacles, primarily by smothering rather than chemical toxic effect, is likely where oil covers rocks; but full recovery is likely to occur by new recruitment in the following year unless residues of oil are persistent (e.g., from viscous oils in sheltered locations). Bleaching of coralline algae (crustose spp. and Corallina spp.), and other red algae, is likely to occur from toxic oil concentrations, but not from weathered oil. However, death of the plants is not inevitable unless oiling and toxicity is very severe, and surviving plants are likely to regain colour quickly. Other algae (e.g., kelp, rock weeds, and foliose red and green algae) appear to be much less sensitive to oil and often show no detectable effect even from heavy oiling, unless there is persistent smothering.

Page 5-6 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Splash zone lichens are vulnerable to oiling on very high tides and some of these long-lived, slow-growing species are sensitive to oil coating their thalli (particularly the fruticose species). Recovery can take many years. The following case studies provide examples of spills with different scales of effect and recovery times: Torrey Canyon: The 1967 Torrey Canyon (Table 5.3) spill is famous for the poor reputation that it gave to the use of chemical dispersants, including their application on heavily oiled exposed rocky shores. The oil itself was relatively much less toxic than the first generation dispersants that were used to combat it. The severe effects on most species (100% of all conspicuous species in some places) that were observed were due to those dispersants. The recovery process of the rocky shore communities is described in detail by Southward and Southward (1978) and Hawkins and Southward (1992). Although the oil and dispersant had almost completely gone from rocky shores within a year, the severity of the acute effect resulted in a protracted recolonization and succession process in some areas, as various species came and went over the course of 3 years. Recolonization of two species on the edge of their geographic range, the limpet Patella depressa and the hermit crab Clibanarius erythropus, was particularly slow and the latter was still not present after 10 years. The lower shore communities recovered relatively quickly, but mid-shore barnacle/limpet/fucoid communities took a number of years. Conditions were considered normal within 6 years, but long-term observations suggested that mid-shore barnacle-limpet-fucoid populations did not stabilize for 10 years, possibly up to 15 years at the worst affected sites. Upper shore lichen populations were also damaged by the dispersant and recovered very slowly.

Table 5.3 Summary Table for the Torrey Canyon Oil Spill Oil Spill Name Torrey Canyon Location , UK Year 1967 Oil Type Light Crude Specific Name Arabian Crude Volume (metric tonnes) 118,000 Platform Tanker Environments Affected Marine Valued Ecological Component (VEC) Studied • Algae • Sediment • Shoreline • Vegetation Summary: On March 18, 1967 the tanker vessel Torrey Canyon grounded on Pollard's Rock on Seven Stones reef. A total of 118,000 metric tonnes of light Arabian crude oil spilled between the Scilly Isles and Cornish mainland. Approximately 50 km of French and 190 km of Cornish (UK) coastline were affected and an oil slick formed measuring 434 km2 The efforts to salvage the spilled oil became useless due to the oil thinning out. Extensive cleanup of affected beaches occurred mainly in the first 6 months after the spill. Studies of algae, shoreline and vegetation were followed sufficiently long for recovery to be evident.

July 2012 Page 5-7

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Arrow: Bunker C fuel oil from the 1970 Arrow spill (Table 5.4) contaminated sheltered rocky shores in Chedabucto Bay, . There was no clean-up of the rocky shores. Although there were no pre-spill data available, comparisons between oiled and unoiled sites suggested that abundance of many species had been reduced by the oil. Abundance of most of those species was back to normal levels within 7 years, although upper-shore fucoids (Fucus spiralis) were still not present on the oiled shores (Thomas 1978). However, whether this species had been present before the spill is uncertain, and it is possible that the lack of recruitment was due to natural environmental factors.

Table 5.4 Summary Table for the Arrow Oil Spill Oil Spill Name Arrow Location Nova Scotia, Canada Year 1970 Oil Type No. 6 Fuel Oil Specific Name Prudhoe Bay Crude, Bunker C Volume (metric tonnes) 7,980 Platform Tanker Environments Affected Marine Valued Ecological Component (VEC) Studied • Fish • Macroinvertebrates • Microbial Community • Sediment • Shoreline • Soil Summary: On February 4, 1970 the Arrow tanker carrying 8,432 metric tonnes of Bunker C fuel oil ran aground on Cerberus Rock inside of Chedabucto Bay in Nova Scotia, Canada. Approximately 7,980 tons of fuel were spilled, oiling 310 km of coastline. Approximately, 48 km of oiled shoreline were heavily cleaned after the spill. The estimated volume of recovery and cleanup end dates could not be found in the literature. Studies of a number of VECs were followed sufficiently for some 30 years after the spill for recovery to be evident. Tsesis: Medium-grade fuel oil and some bunker fuel from the 1977 Tsesis spill (Table 5.5) contaminated sheltered rocky shores in the Swedish sector of the Baltic Sea. The fucoid algae that dominated the shores were unaffected, but the animals within the algae, including bivalves, snails, and small crustacea, were severely diminished in the worst affected areas. Recovery started within 2 months and densities had returned to pre-spill levels after one year at some sites. Recovery at the worst-affected sites was not complete when the monitoring terminated but was expected to be complete within three years (Linden et al. 1979).

Page 5-8 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Table 5.5 Summary Table for the Tsesis Oil Spill Oil Spill Name Tsesis Location Baltic Year 1977 Oil Type No. 5 Fuel Oil Specific Name medium grade fuel oil (11) Volume (metric tonnes) 1,000 Platform Tanker Environments Affected Marine Valued Ecological Component (VEC) Studied • Algae • Macroinvertebrates • Shoreline Summary: On October 26, 1977 the tanker Tsesis grounded and spilled approximately 1,000 metric tonnes of fuel oil while entering the Sodertalje channel south of Stockholm, . The initial cleanup of some 18 km of shoreline went on for weeks after the spill. The estimated volume of recovery was between 600 to 700 metric tonnes. Studies of plankton, macroinvertebrates and shorelines were followed to address the progress of recovery. Esso Bernicia: Heavy fuel oil from the 1979 Esso Bernicia spill contaminated sheltered rocky shores in Sullom Voe, Shetland Islands. The shores varied from steep bedrock to gradually sloping boulder/cobble shores. Oil and oiled substrata were removed from some of the latter using bulldozers. A number of pre-established rocky shore monitoring sites had been surveyed the year before the spill and new sites were quickly established to monitor the recovery of the bulldozed sites. After the response, the remaining oil cover decreased fairly rapidly at first, particularly on the lower and middle shore, but then more slowly over many years. At sites that were not cleaned, some effects on lichens, limpets, and snails were recorded, but recovery was rapid. There were no detectable effects to the monitored communities at those sites within a year. Any effects were less than the level of natural fluctuations. This was not the case at the bulldozed shores, where continued instability of the substrata was evident for many years, with consequent slow recolonization of species that require greater stability. However, the substrata eventually re-found its equilibrium and full recovery of those communities occurred within 15 years (Moore et al. 1995). Some patches of weathered tar are still present in upper shore zones at a few sites, and thin silvery sheens can occasionally be seen in sheltered upper shore rock pools at one such site. These occasional patches of weathered tar have no apparent effect on the immediately associated upper shore species, including large numbers of littorinid snails in the tar-contaminated pools and apparently healthy algae growing on asphalt pavement (personal observations by J. Moore). Exxon Valdez: The 1989 Exxon Valdez spill resulted in oiling of over approximately 2000 km of coast, most comprising mixed bedrock-boulder-cobble-gravel shores, ranging from very exposed to moderately sheltered. Intensive and extensive clean-up was carried out, including warm-water flushing. Natural forces continued to remove oil and after 3 years the estimated length of oiled

July 2012 Page 5-9

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

shoreline in Prince William Sound was 10 km (approximately 0.2% of the total PWS shoreline) (Neff et al. 1995). However, some oil persisted in places where it had accumulated under the surface of the coarse substrata and was protected from the cleaning action of strong water movement by rock armouring (boulders). Michel et al. (2010) estimated that 97 tonnes of weathered oil (approximately 0.2% of the original spill volume) were still present in Prince William Sound in 2007 and that it would be many more decades before it was all gone. Sediment bioassay tests have suggested that the residual oil does have remaining toxicity (Day 2005) although methodological issues confuse interpretation. A number of studies have demonstrated evidence of exposure effects (using biochemical biomarkers) in some fish and birds, but there is little conclusive evidence to suggest that the residual oil is influencing community succession or mature benthic communities (Integral Consulting 2006). Monitoring on treated and untreated rocky shores, was carried out between 1989 and 1997 (Houghton et al. 1997, followed by Coats et al. 1999). Houghton et al. showed that the communities on untreated sites had mostly recovered after two years, but that treated sites had not. Warm water treatment resulted in losses of natural rocky shore fauna and flora, including the majority of the rockweed. Recolonization and growth of all the typical species occurred gradually over the first three years, by which time much of the mid and upper shores was densely covered in rockweed plants, all of a similar age. However, many of these plants then died two years later, presumed to be at the end of their natural life, with consequent loss of associated animals. Another cohort of rockweed sporelings replaced them the following year and started to grow, accompanied by gradual recolonization of the associated animals. It was presumed that this pronounced cyclical change in the rockweed cover, akin to that described by Southward and Southward (1978) following the Amoco Cadiz spill would continue until a more natural mixed age class structure was re-established. Some authors suggested that could take many years (Integral Consulting 2006), but no survey data are available since 1997. Houghton et al. (1997) highlighted the contrast between their conclusions and those of Gillfillan et al. (1995), who concluded that the shorelines of Prince William Sound ‘had largely recovered from the effects of the spill by the summer of 1990’. The difference was due to differences in survey design: the Houghton et al. design selected the worst affected sites to compare with reference sites and collected considerably more sample data per site than Gilfillan et al. (1995), enabling much greater statistical power to detect differences. Gilfillan et al. carried out a broad- scale random sampling design and less intensive sampling, which could not detect the effects clearly shown by the Houghton et al., but better represented the scale of the effects to the whole Prince William Sound system. Further, Coats et al. (1999) and Skalski et al. (2001) consider the real possibility that the reference sites and affected sites were different before the spill, that natural fluctuations could explain the community differences and that recovery had already occurred. It is certainly likely that the scale of the oscillations and cyclical changes observed at the treated sites, after the first three years, can, and often do, occur quite naturally. So even if the oscillations and cyclical changes at the treated sites happened to be unnatural, their effect on the ecosystem will have not been ecologically meaningful at even a relatively local scale.

Page 5-10 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Sea Empress: Forties blend crude oil from the 1996 Sea Empress spill (Table 5.6) contaminated approximately 200km of shoreline; half of that was classified as heavily oiled and approximately half of that was exposed rocky shore. Only a very small proportion of that rocky shore received any clean-up attention because it was known that the oil would be removed rapidly by natural forces. A number of pre-established rocky shore monitoring sites were present in the region and many other sites were also studied after the spill. Severe acute effects were found at sites contaminated by the freshest, most toxic, oil, but effects were much less obvious at others and undetectable at some which did not receive the oil until it was well weathered (at sea for over a week). Natural removal of the oil was as rapid as expected, with almost no visible oil remaining on exposed rocky shores after 1 year. Studies at the most severely affected sites showed that the rocky shore communities were similar to pre-spill communities, with no detectable loss of biodiversity or other effects, within 5 years. Most of that recovery took place within 2 years. On less severely affected shores, recovery was complete in less time. Effects on splash zone lichen communities were described at some sites, particularly at one location where inappropriate pressure washing stripped lichens from the rock and lichen cover is still very low (Crump et al. 2003; Moore 2006b; personal observations by J. Moore).

Table 5.6 Summary Table for the Sea Empress Oil Spill Oil Spill Name Sea Empress Location , UK Year 1996 Oil Type Crude / No. 6 Fuel Specific Name none Volume (metric tonnes) 72,480 Platform Tanker Environments Affected Marine Valued Ecological Component (VEC) Studied • Algae • Macroinvertebrates • Shoreline Summary: On February 15, 1996 the tanker vessel Sea Empress became grounded on mid-channel rocks at St. Ann's Head at the entrance to Milford Haven Waterway in Pembrokeshire, Wales. Over 72,000 metric tonnes of oil spilled into the ocean with the remaining 57,000 on board being pumped out for refining. Approximately, 200 km of coastline was affected. The estimated volume of recovery from the initial cleanup efforts was 2 percent. Extensive cleanup efforts began right after the spill and went on for several weeks, while reduced scaled cleanup efforts went on for an entire year. Studies of algae, macroinvertebrates and shorelines were followed sufficiently long for recovery to be evident. Recovery of a well-studied population of the rarely recorded cushion starfish (Asterina phylactica) was much faster than initially expected. Mortality of the cushion stars was very high (greater than 95%) because their rockpools were severely oiled. Recovery of the population seemed unlikely because they are long lived and normally reproduce sexually and brood their young in situ (therefore no recruitment from planktonic larvae). However, a return to pre-spill densities occurred within 6 years. During this period it was discovered that this species is capable

July 2012 Page 5-11

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments of self-fertilization (parthenogenesis), allowing the remaining isolated animals to build up a viable population and then continue normal sexual reproduction (Robin Crump, pers. comm.). AURIS (1994) summaries the recovery processes from a number of other oil spills on exposed or moderately exposed rocky shores, some with very rapid recovery, some taking longer; but the case studies given above describe much of the range.

5.1.2.2 Sediment Shores There are number of physical and biological characteristics of sediment shores that can influence the effects of oil spills and the rates of recovery; including wave exposure, shore topography, sediment composition, height of water table, presence of large burrows, abundance and diversity of infauna, and use of the shore by birds for feeding and roosting. Wave exposed clean sandy shores are often considered to have a low sensitivity due to the natural cleaning of the waves and the relatively poor fauna in the sediment. However, a sheltered muddy gravel shore with a high biodiversity including numerous long-lived bivalves, would have a higher sensitivity. Oil can persist and remain toxic in sheltered muddy sediments for many years (sometimes decades), particularly in anoxic sediments. AURIS (1985) collated information on effects and recovery rates in intertidal sediments from a large number of spills and experimental studies. IPIECA (1999) summarizes information on their sensitivity and recovery potential. As on all shores, but particularly on wave-sheltered shores, bulk oil tends to concentrate along the strandline. As a result, hydrocarbon contamination in lower and middle shore sediments is usually less conspicuous and less persistent. However, if the oil is very fresh and toxic and/or water column concentrations are high, any sensitive fauna may be severely affected by the acute exposure. Filter feeding animals, like most bivalves and some that amphipods, which live in the sediment but take their food from the water above the sediment surface, are particularly susceptible to toxic concentrations in the water. Acute effects can occur even if the sediment remains uncontaminated. Recovery time after such effects is then a function of the recruitment potential and growth rates. Where oil is persistent, recovery of the associated sediment fauna is usually correlated to the rate of degradation of the most toxic components. This is largely a function of water movement but can be slow if the oil has penetrated deep below the surface. Penetration of oil into intertidal sediments depends on the sediment composition and the extent to which the sediment animals and plants create burrows and channels. Interstitial water movements in sands and gravels allow deeper penetration of oil, but this movement also aids natural cleaning. Muddy sediments resist such penetration. Large crab burrows are less common in intertidal sediments of temperate and subarctic zones than in warmer zones, but the smaller burrows of other macrofauna and the dead stems of saltmarsh plants can allow penetration of low-viscosity oils. Long-term physical smothering by tar mats can occur on sediment shores, but unlike in saltmarshes, spilled oil rarely persists on such sediment shores because it is usually easy to remove it or break it up without additional physical effects.

Page 5-12 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

The following case studies provide examples of differing scales of effects and recovery times for sediments: Arrow: Bunker C fuel oil from the 1970 Arrow spill contaminated (Table 5.4) a very sheltered muddy bay in Chedabucto Bay, Nova Scotia. MacDonald and Thomas (1982) describe the results of monitoring over nine years. Surveys in the first year after the spill showed that there were heavy mortalities of the bivalve Mya arenaria, and the follow-up surveys showed that recruitment was greatly affected by persistent contamination, but improved as oil toxicity declined. Six years after the spill, toxic levels of oil still remained in some locations and analysis of M. arenaria growth rates (from length and weight frequency data) from four oiled and four unoiled sites showed that growth rates were significantly reduced at oiled sites. Lee et al. (1999) carried out bioassay studies in more recent years (1993 and 1997) on sediments from the same area. They showed that sediments from the oiled sites had low toxicity, as measured by exposure effects indicators in flat fish (mixed function oxygenase enzyme systems) and other toxicity tests (including amphipod survival tests). Sea Empress: Forties blend crude oil from the 1996 Sea Empress spill (Table 5.6) resulted in heavy oiling of a variety of sediment shores in Milford Haven, UK, ranging from wave-exposed sand beaches to sheltered muddy sand flats. Pre-spill data from two years earlier were available for a number of the oiled shores and additional monitoring sites were established within a month after the spill. Hydrocarbon analyses showed that sediment concentrations in some locations were initially high, but reduced rapidly and were relatively low, even in muddy sediments (though still elevated above typical pre-spill levels) within 18 months. A program of sediment macrofauna sampling on selected sandy beaches and muddy shores was initiated immediately following the spill and continued, initially monthly then less frequently, for 18 months. Initial results showed reductions in numbers of amphipods and some molluscs, particularly cockles, and increases in some opportunistic polychaetes. However, within one year the populations of most species were back to typical pre-spill levels, as was species diversity (Rostron 1998). Large fluctuations in some species densities were still occurring but the levels of natural fluctuations were not known and were often high. Repeat surveys of meiofaunal communities in sandy shores, nine months after the spill found no effects of the spill that could be distinguished from the natural fluctuations (Moore et al. 1997). It was considered likely that an earlier survey (within a few weeks of the spill) would have detected at least some gross effects; but recovery was apparently rapid. Exxon Valdez: Studies on the effects of intensive shoreline clean-up on lower shore sediment clam populations found reduced abundances and other significant effects, due to disruption of the surface armouring, that were still evident 13 years later (Lees and Driskell 2007). Results from repeat surveys in 2010 were inconclusive due to large changes in the physical characteristics of the shoreline at all sites probably unrelated to the treatments (Lees 2011). The PWS intertidal fauna has been observed to respond to interannual shifts in weather conditions and ocean climate (Lindstrom et al. 1999).

July 2012 Page 5-13

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

5.1.2.3 Saltmarshes Long-term effects from oil spills are more likely in saltmarshes than in any other intertidal habitats of cold temperate and sub-arctic regions, due to the persistence of any oil accumulations and the potential for severe physical damage during clean-up. There are a number of case studies that describe such long-term effects on some saltmarshes where oil still persists after more than a decade. However, in the majority of cases, full recovery of the marsh communities does occur, often in less than five years and usually in less than ten years. Further, in line with all other resources affected by oil spills, recovery processes are normally rapid once initial recolonization has taken place, and slow down only in the later stages. Long-term effects are therefore normally subtle (i.e., difficult to detect within the levels of natural fluctuation) and often of limited spatial extent. Analysis of the numerous case studies have shown that the effects and recovery rates of oiled saltmarshes can vary considerably and that even heavy oiling does not necessarily result in large scale mortality of saltmarsh plants or slow recovery. The factors that affect the scale of effect and recovery time have been described in IPIECA (1991) and Baker et al. (1996). The two main factors are 1) oil type, amount and degrees of weathering and 2) the season of the year and the species present during that season. Lighter and more penetrating oils are more likely to cause acute toxic damage than heavy or weathered oils. The thickness of the oil deposit and the degree of penetration will also affect long-term persistence, smothering of the soil surface, and the rate at which degradation processes reduce toxicity. Oil deposits are typically very patchy so that recovery of the marsh occurs at different rates in different areas. Concerning seasonality, the underground systems of perennial species, which often dominate the mid- and upper zones of marshes, are relatively protected from oil contamination. During the winter, they store food reserves, which are then gradually depleted over the course of the spring and summer as the above ground vegetation grows. Loss of above-ground vegetation will therefore be much more serious in the late summer and early autumn, but will have little effect in the winter or early spring. Annual species have less well-developed underground systems, and thus are more vulnerable in the summer. Seeds and seedlings of any species will be most vulnerable in the spring. Baker et al. (1996) went on to describe a number of scenarios that have been encountered following oil spills and field experiments; they are summarized as follows: • Light to moderate oiling, oil mainly on perennial vegetation with little penetration of sediment. Some or all of the plant shoots may be killed, but recovery can usually take place from the underground systems. Good recovery commonly occurs within one to two years. • Light to moderate oiling, oil mainly on annual vegetation with little penetration of sediment. It is possible that areas of vegetation may die completely. If large areas are affected, recovery may be delayed because seed has not been produced or cannot germinate because it has been oiled.

Page 5-14 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

• Oiling of perennial vegetation such that species composition is altered. Following oiling, it is sometimes found that species composition is altered for some time because relatively resistant species take over from more susceptible species. Provided a good vegetative cover (of whatever species composition) is established quickly, there will be minimal risk of soil . • Oiling of shoots combined with substantial penetration of oil into sediments. This is more likely to happen with relatively fresh light crude oils or light refined products, such as No. 2 fuel oil [diesel], because these are less viscous. Damage to the underground systems results from the sub-surface oil, and recovery is delayed. Areas of vegetation may die completely. Sediment erosion may occur if recolonization does not start within a year. • Thick deposits of viscous oil or mousse on the marsh surface. Vegetation is likely to be killed by smothering, and recovery delayed because persistent deposits inhibit recolonization The above discussion focused on effects to and recovery of the saltmarsh vegetation. Most studies have found that effects and recovery of associated fauna (including birds, reptiles, crabs and sediment infauna) are closely related to the vegetation. However, the adequacy of vegetation attributes as indicators of recovery of all marsh components will depend on the quality of the vegetation study methods. Some studies have described toxic effects on saltmarsh sediment fauna in marshes that have no visible signs of effect to the vegetation cover, although more- detailed vegetation studies may have detected effects. The following case studies provide examples of differing scales of effect and recovery times: Florida: No. 2 fuel oil (heating oil) from the 1969 Florida spill (Table 5.7), contaminated saltmarshes in West Falmouth region of Buzzards Bay, . Toxic effects on fiddler crabs within the marsh were monitored for 7 years and showed gradual recovery, which was correlated to with the loss of PAHs from the sediment, but the crabs still continued to avoid sediments where high petroleum concentrations remained (Krebs and Burns 1977). Oil was present under the surface in some marsh locations after 20 years and slight exposure effects (biomarkers) were found in marsh fish (Teal et al. 1992). The same location was sampled again 30 years after the spill and the remaining oil was in a band, 6 cm below the bottom surface. Despite the subsurface oil residues, the marsh at the site was healthy (Reddy et al. 2002).

Table 5.7 Summary Table for the Florida Oil Spill Oil Spill Name Florida Location Falmouth, Massachusetts Year 1969 Oil Type No. 2 Fuel Specific Name none Volume (metric tonnes) 557 Platform Barge Environments Affected Marine

July 2012 Page 5-15

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Table 5.7 Summary Table for the Florida Oil Spill (cont’d) Valued Ecological Component (VEC) Studied • Macroinvertebrates • Shoreline Summary: On September 16, 1969 the oil barge Florida ran aground in Buzzards Bay (Cape Cod) off the coast of West Falmouth, MA USA and spilled over 500 metric tonnes of No. 2 Fuel oil. There were no details found regarding cleanup efforts and oil recovered in the current literature. Core samples in the salt marshes that were taken 40 years after the spill still show oil present in the sediment. Milford Haven: Heavy fuel oil from a 1969 spill in Milford Haven, UK, heavily fouled an area of saltmarsh. Complete recovery of the marsh vegetation occurred within 15 years, although a well-defined layer of oil was still present under the sediment surface in part of the marsh three years later. Analysis of the oil layer after 16 years showed that it was relatively unweathered, with a high aromatic content, but live plants were growing below, through and above the layer (Baker et al. 1993). Metula: Arabian crude and bunker C fuel oil from the 1974 Metula spill heavily oiled areas of saltmarsh in the , . The thickness of the oiling and the cold climate resulted in very slow recolonization of several hectares of marsh, and consequently, bare areas were still present within the marsh after 23 years (Owens et al. 1999). Recolonization and recovery had occurred in many parts of the marsh where smothering had been limited and where some remediation trials (physical tilling / mixing) had been conducted, but without further intervention it is expected that full recovery will take many decades (Owens and Sergy 2005). Amoco Cadiz: Large areas of the Isle Grande marshes in were heavily oiled by the 1978 Amoco Cadiz spill. Some areas were intensively cleaned using machinery and large numbers of soldiers, while other areas were untreated. It was later realized that removal of marsh sediment caused serious damage to the marsh by changing its height within the intertidal zone. Pre-spill data were limited to aerial imagery, but vegetation studies that monitored the post-spill changes of oiled and unoiled marshes and treated and untreated marshes provided a good description of the recovery processes (Baca et al. 1987). In the untreated marsh, it was shown that although heavily oiled, many of the perennial plants survived because their below-ground parts were unaffected. This aided recolonization of more severely affected areas. Recolonization of annuals, followed by perennials, occurred during the first 3 years. Recruitment of marsh fauna then occurred very rapidly and the untreated marsh was considered to have recovered within five years. In the treated marsh, recovery of the normal marsh species was delayed by initial colonization of opportunist species, which responded rapidly to the disturbance created by the remediation attempts. However, vegetation cover increased over the course of six years, as did the recruitment of marsh fauna. After eight years, the marsh vegetation was still extensively altered and some bare ground was still present, even with the aid of restoration planting, but recovery was considered to be in its last stages.

Page 5-16 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Table 5.8 Summary Table for the Oil Spill Name Amoco Cadiz Location Year 1978 Oil Type Light Crude Specific Name light Arabian, light Iranian Volume (metric tonnes) 240,000 Platform Tanker Environments Affected Marine Valued Ecological Component (VEC) Studied • Algae • Fish • Macroinvertebrates • Shoreline • Vegetation Summary: On March 18, 1978 the tanker vessel Amoco Cadiz grounded on Portsall Rocks and spilled approximately 240,000 metric tonnes of light crude oil 5 km off the coast of Brittany, France. The entire cargo including crude and fuel oil spilled into the sea with approximately 320 km of affected coastline. The estimated volume of recovery from the initial cleanup efforts was 100,000 metric tonnes. Extensive cleanup of affected areas occurred mainly 6 months after the spill. Studies of a number of VECs were followed sufficiently long for recovery to be evident. Sea Empress: Forties blend crude oil from 1996 Sea Empress contaminated a number of small saltmarsh areas in Milford Haven, UK, and was observed to cause some dieback of the vegetation where oiling had been substantial (SEEEC 1998). A series of surveys to assess the effect of the spill on these saltmarshes (Bell et al. 1999) suggested that there had been some effects to vegetation in the marshes that were coated with oil, but detection was limited by lack of pre-spill data and the late start of the study. Much of the vegetation that had been oiled had survived with no apparent damage but there was a patchy dieback of vegetation in some marshes. A re-survey one year later found good recovery of most species in most locations, and re-visits after six years found no differences between sites affected by the spill and those either protected from its effects or situated beyond its zone of impact (Prosser and Wallace 2003). They concluded that the saltmarsh vegetation was no longer influenced by the effects of the spill.

5.1.3 Plankton The literature on the effects of oil on plankton is mostly based on laboratory studies, where planktonic organisms have been exposed to fractions of oil in controlled conditions. There have also been a few studies based in larger enclosed ‘mesocosm’ systems. Such studies have described a wide range of acute, chronic and sublethal effects to various species, including adults, juveniles, and eggs. Eggs and juveniles, including larvae of fish and benthic invertebrates, are usually much more sensitive than are adults. However, while laboratory and mesocosm studies have provided valuable information on toxicity and effects of various hydrocarbon

July 2012 Page 5-17

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

concentrations, the fact that the experimental systems are enclosed does not allow natural recovery processes to take place. Planktonic organisms are clearly very sensitive to toxic hydrocarbons, but developing an understanding of the scale of the effect of an oil spill on real-world plankton populations and communities and the rate of recovery from that effect requires field studies in the open sea. Methods for studying the effect of oil concentrations that cannot be seen, on organisms that cannot be seen, in a medium that is constantly moving are not straightforward. The levels of natural patchiness and temporal fluctuations in plankton communities are also very high. Nevertheless, a number of field studies been carried out after accidental oil spills and some have detected acute effects, but only very transitory. Recovery is very rapid, due to the wide distribution and rapid regeneration rates of most species (Wells 1985, Michael 1977). In enclosed waters, such as lagoons or shallow inlets limited recruitment from outside could theoretically slow recovery. The main literature review sources used in this section are Davenport (1982), Wells and Percy (1985), National Academy of Sciences (1985 and 2003) and AMAP (2008). The following case studies are available: Exxon Valdez: Zooplankton is the main food source of many juvenile marine fish and assessing the effect of the oil spill on zooplankton biomass provides a straightforward approach to ascertaining potential effects of the spill on the food. Such data have been collected by the Prince William Sound Aquaculture Corporation (PWSAC) since the late 1970s for use in synchronizing hatchery fry releases in the Sound with the plankton bloom (Cooney et al. 1981) and provides a relative index for year-to-year comparisons. The sites monitored are Port Ashton and Elrington Passage, both within the oil spill impact area. Large calanoid copepods, dominated by Pseudocalanus spp. and Calanus spp., are the primary prey of young salmon, and crucial to marine productivity. Data from monitoring sites, representing the top 20 m of the water column, indicated that zooplankton biomass in 1989 was among the highest recorded in western PWS (2 x the previous 8 year mean), and ostensibly was responsible for exceptional growth among juvenile salmon for that year. These data also correspond with the data on plankton biomass gathered by Celewycz and Wertheimer (1996), showing no differences between oiled and non-oiled areas in any of the eight sample sites except for greater abundance of some species in oiled areas. The distribution and density of epibenthic harpacticoid copepods was also monitored in heavily and lightly oiled bays by Wertheimer et al. (1996). They found that copepod mean densities on heavily oiled shorelines were higher than on lightly oiled shorelines. Moreover, the percentage of egg-bearing copepods were similar in the oiled and non-oiled sample sites. Tsesis: When 1000 tonnes of medium grade fuel oil was spilled into the Baltic in 1977 from the Tsesis, scientists from a nearby marine ecology station responded rapidly to study possible effects on the plankton (Johansson et al. 1980). They showed that zooplankton biomass had declined substantially close to the wreck, during the first few days after the spill, but that it was re-established within 5 days.

Page 5-18 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Amoco Cadiz: Severe oiling of the northwest coast of France following the 1978 Amoco Cadiz spill resulted in an almost complete loss of plankton from two shallow embayments (‘abers’) for a month afterwards. The plankton recovered quickly and became abundant in one of the estuaries after 3 weeks. The spring plankton bloom appeared to be depressed but after three months, there were no detectable effects (Spooner 1978, Simian et al. 1980). Sea Empress: Some studies of the plankton communities around the west coast of Wales were carried out following the 1996 Sea Empress spill (SEEEC 1998 and Batten et al. 1998). They found no detectable effects, although there was a temporary reduction in abundance of barnacle larvae in spring 1996. However, effects of that reduction did not limit the settlement and recruitment of barnacles on the adjacent coastline, which was particularly strong that year.

5.1.4 Benthic Biota Seabed communities can be directly exposed to oil from spills through a number of physical processes; the nature of which then affects recovery processes (Lee and Page 1997): • sinking of heavy oil residues – occurs only in unusual circumstances where the oil becomes heavier than seawater (e.g., after partial burning of heavy fuel oil). Such residues can be very persistent and could potentially smother seabed habitats, but are more often present as smaller particles. (i.e., tarballs). • mixing of sandy shoreline sediments and oil, to create a material that is heavier than seawater and settles on the seabed close to the beach – has been described from a few spills on wave exposed sand beaches, but not commonly. The oil/sand mixture can form persistent mats in the shallows, which gradually break up and create persistent tarballs. • oil in water, either fine droplets of dispersed oil (physically or chemically dispersed) or water-soluble fractions, in the absence of suspended sediments, mixed down into water column where they make contact with seabed organisms (e.g., filter feeders) – likely to occur only in turbulent shallow water (<10m). In the absence of suspended sediment (silt/clay particles), the oil is unlikely to contaminate the sediment. Recovery of seabed life that was acutely affected by the toxic concentrations should occur without any hindrance from persistent oil. • dispersed oil droplets adsorbed to suspended sediment making them heavy enough to settle to the seabed – likely to occur to some extent in any coastal situation where there is suspended sediment (silt/clay particles). Studies suggest that the deposition is typically spread over a large area of seabed (not just in shallows), resulting in a thin layer of contaminated flocculated fine particles (‘clay-oil flocs’) on the sediment surface. The hydrocarbons in this floc will degrade rapidly and may not significantly increase the concentrations above background levels, even in the surface sediments. Seabed organisms, particularly filter feeders, may be acutely affected by the hydrocarbons in the floc, but their recovery should occur without any hindrance from persistent oil.

July 2012 Page 5-19

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

As with the previous example, but in locations where the suspended sediment load and the amount of oil in the water are both particularly high, can result in heavy deposition of contaminated particles which become incorporated into muddy seabed areas (‘sediment sinks’), where they can persist for many years. Toxicity of the seabed oil will depend on the oil type, the concentration, and the degree of weathering that occurred before settling out. A review of oil spill effects in the sub-tidal by Lee and Page (1997) highlights the typically much lower concentrations of oil found in sub-tidal sediments compared to those in the intertidal. Of the five processes described above, only the third and fourth occur commonly, and they do not result in persistent seabed concentrations. Nevertheless, high concentrations have occurred, notably from the Amoco Cadiz and Braer spills. Some groups of sediment fauna are more sensitive to oil than are others. Amphipods (particularly the filter-feeding tube-dwelling species, e.g., Ampelisca spp.), filter feeding bivalves and burrowing urchins (e.g., sand dollars) have been identified as the main casualties in a number of oil spills. Large numbers of filter-feeding bivalves and burrowing urchins are often washed up on beaches after spills. As noted above, recovery of acutely affected seabed communities that are not contaminated by hydrocarbons will occur through natural processes, and the rate of recovery will depend on the rate of recruitment and longevity of the affected species. Where oil contaminates the sediment, the rate of recovery will depend on the loss of toxicity. When hydrocarbon concentrations become are low enough, the sediment becomes suitable for opportunistic species that can take advantage of increases in the microflora feeding on hydrocarbons (Lee and Page 1997). As the hydrocarbons become further degraded, the densities of opportunistic species decline, and the normal suite of species returns. The following case studies provide examples of differing scales of effect and recovery times: Florida: No. 2 fuel oil (heating oil) from the 1969 Florida spill, contaminated shallow sub-tidal sediments in West Falmouth region of Buzzards Bay, Massachusetts. The natural infaunal community was replaced by opportunist polychaete worms for many months until the toxicity of the sediment declined and the natural community was re-established. Biodiversity had not fully returned after five years, but recovery was progressing well (Michael et al. 1975). Tsesis: Medium grade fuel oil from the 1977 Tsesis spill was naturally dispersed into the Baltic waters where it mixed with suspended particles and contaminated seabed sediments at depths below 30 m. Notable reductions occurred in populations of benthic amphipods and other species of macrofauna and meiofauna, but recolonization of affected species was evident in the first two years after the spill. Recolonization was slower at the most substantially affected sites, but the community then became dominated by a bivalve that replaced the amphipods. Elmgren et al. (1983) considered that this switch in the community was an effect of the spill and that the recovery could not be considered to have occurred until the amphipod community had restored itself. Others (e.g., Kingston 2002) argued that as productivity and biodiversity had been re- established much earlier, then recovery could be considered complete.

Page 5-20 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Amoco Cadiz: High hydrocarbon concentrations were found in sub-tidal sediments of some French estuaries following the 1978 Amoco Cadiz spill. These resulted from very high suspended sediment loads in the estuary waters together with natural mixing of large quantities of oil. Opportunistic polychaete worms became dominant for a few years until the sediment toxicity had declined (Glemarec and Hussenot 1981). Further along the coast, in the Bay of Morlaix, two sediment communities responded very differently: a muddy, fine-sand community, which became contaminated by hydrocarbons, showed relatively little consequence and recovered very quickly (Dauvin 2000), while a fine-sand community, which was not significantly contaminated, was severely affected and took more than 10 years to recover. The difference was due to the presence of high densities of the amphipod Ampelisca spp. in the fine-sand community, which was dramatically reduced over large areas of seabed. Although the amphipod has a short life span and breeds more than once per year, recolonization and establishment of the dense ‘climax community’ occurred slowly (Dauvin 1998). Nella Dan: Diesel spill from the 1987 Nella Dan spill in the Australian sector of the Antarctic resulted in high mortalities in invertebrate communities of intertidal and shallow sub-tidal rock habitats. The intertidal communities recovered quickly, but kelp holdfast communities showed longer-term effects. Seven years after the spill, the holdfast community structure in samples from heavily oiled sites showed moderate levels of recovery, with increased abundances of sensitive species. However, holdfasts at sites that were filled with sediment containing traces of diesel oil were dominated by opportunistic worms. Exxon Valdez: Sub-tidal sediment hydrocarbon concentrations in nearshore areas of Prince William Sound following the 1989 Exxon Valdez spill were initially elevated, but the amounts were mostly very low and they rapidly returned to background levels in most areas (Short et al. 2003). However, toxic concentrations were found in muddy sediments associated with eelgrass beds in a few shallow bays that received heavy oiling. Concentrations of PAHs had declined by 90% within six years. Effects on the fauna within the contaminated sediment included reductions in populations of species known to be sensitive to toxic hydrocarbons and increases in typical opportunistic species, but the communities had mostly recovered within six years. Some differences between oiled and reference sites remained, but it was uncertain whether were this was due to inherent differences between the sites or effects of the spill (Jewett et al. 1999). There were higher densities of some kelp and lower densities of some large epifauna, at oiled sites compared to reference sites one year after the spill, but the differences were no longer evident within one for the kelp and four years for the epifauna (Dean et al. 1996a,b). Differences of opinion between researchers about biological effects in the subtidal from the residual oil, based mainly on exposure effects (biomarker) studies, continued for many years (Jewett et al. 2003). However, the absence of elevated PAHs in mussels in 2005 (Page et al. 2005) strongly indicated that any remaining oil was no longer bioavailable. The Exxon Valdez Oil Spill Trustee Council considers that remaining Exxon Valdez oil in sub-tidal sediment is no longer a concern and that sub-tidal communities have Very Likely Recovered (EVOS Trustee Council 2010, Integral Consulting 2006).

July 2012 Page 5-21

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Braer: The 1993 Braer spill occurred during violent storms on a very exposed coastline, so most of the light Gulfaks crude oil was rapidly dispersed into the water column, resulting in extremely high concentrations of dispersed oil near the wreck for a few days, together with fine particles of silt and clay. This resulted in contamination of offshore seabed sediments in deep water, with concentrations reaching 10,000 ppm. Despite the high concentrations, the infaunal communities showed relatively limited effects, except for loss of amphipods where oil concentrations were highest (Kingston et al. 1997). Monitoring of oil concentrations showed a 100 fold decrease over ten years (FRS/MS report). Studies of exposure effects (Cytochrome P4501A enzyme induction) in flatfish (Stagg et al. 1998) found evidence of exposure from the most contaminated area in the first year after the spill, but not a year later. Sea Empress: Surveys of seabed sediment communities around the coast of southwest Wales were carried out for number of years following the 1996 Sea Empress spill, and compared with pre-spill data from two years before the spill (Rutt et al. 1998). They showed marked reductions in densities of small crustaceans (amphipods and cumaceans, particularly Ampelisca spp) near the wreck, but no other notable effects on the macrofauna and no evidence of sustained contamination. Monitoring studies showed a clear pattern of recovery of Ampelisca spp. and other amphipods over a period of five years, with densities similar to pre-spill levels by the year 2000 (Nikitik and Robinson 2003). Conspicuous effects of the oil spill, and of the enhanced oil dispersion by chemical dispersant spraying, were large numbers of stranded bivalves and heart urchins (SEEEC 1998). The potential effect on populations of these species in shallow sub-tidal areas of Carmarthen Bay initiated a number of follow-up studies. Apart from the effect on seabed communities, the consequential effect on predators, particularly common scoter, if food availability was reduced was also a concern. Surveys of the benthic communities of Carmarthen Bay did not find any detectable effects of the oil spill on the seabed communities, and populations of the shallow- water bivalves appeared healthy and in high densities (Woolmer et al. in press).

5.1.5 Marine Fish Although concern for the effects on fish quickly arise following a spill, mass mortalities of fish following marine oil spills are rare. Where mass mortalities are reported, it is for spills, such as the Amoco Cadiz, during which turbulence from waves and currents physically dispersed oil into the water column. In the Braer spill, high concentrations of oil occurred in the water column during severe weather and were associated with elevated levels of CYP1A and EROD activity in caged Atlantic salmon at the open water fish farm in a highly affected area (Stagg et al. 1998). These biomarker levels indicate exposure to the spilled oil and decreased to background within 8 weeks. For demersal fish in spills such as Amoco Cadiz and Braer, where dispersed oil has contacted sand and led to high levels of hydrocarbons in bottom sediments, exposure to oil lasts longer than for pelagic fish (Conan 1982; Stagg et al. 1998). For dab, a flatfish, biomarker levels took slightly over a year after the Braer spill to return to background levels (Stagg et al. 1998). For several fish species, frequencies of fin rot disease following the Amoco Cadiz spill decreased from about 80% to “nil” in 3 years. Herring are an important commercial species with a life

Page 5-22 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments history that makes them vulnerable to effects from oil spills and have received sufficient long- term study to assess the factors governing recovery, and are discussed below.

5.1.5.1 Herring Pacific herring, Clupea pallasii, are finfish of great ecological, economic, and social value. As a key-stone forage fish, herring provide a heavy link in the food web between its zooplankton prey and its many predators, such as, salmon and other fish, marine birds, seals, sea lions, and humpback whales (Spies 2007). On the West Coast of North America, herring roe and herring egg-on-kelp are an important traditional food and are sought in valuable commercial fisheries from the Bering Sea through British Columbia and south to San Francisco. Whole herring are also sought as bait for other commercial fisheries as well as recreational fisheries. In several locales, herring provide the first fishing opportunity after the winter season. Subsistence use of herring also confers social and economic benefits. Pacific herring have a complex life history (Figure 5.1, Hay 1985, Northcross et al. 2002). The adults are pelagic and move from offshore summer feeding areas to inshore overwintering grounds. In the spring, adult herring aggregate near spawning grounds. When the gonads mature, the ripe herring move into shallow water to spawn. Spawning females deposit eggs on seaweed, eelgrass, and kelp in intertidal and shallow sub-tidal areas and then spawning males broadcast sperm (milt) into the water over the eggs to fertilize them. In Prince William Sound, the eggs incubate for about 20 days until hatching. After hatching, the yolk-sac larvae remain near the spawning areas until the yolk sac has been absorbed. The older larvae then join the drift of plankton that can carry the larvae some distance from the point of hatching. Depending on temperature and feeding conditions, the larvae metamorphose into young of the year (YOY) herring about three months after hatch. Age 0 and Age 1 juvenile herring grow in the nearshore areas of sheltered bays until they begin to form schools at two years of age. Herring recruit to the adult spawning population at 2 to 5 years of age depending on latitude. Herring in California, British Columbia, and Prince William Sound recruit at ages 2, 2 to 3, and 3 to 5, respectively. The life history habit of spawning in the intertidal and shallow sub-tidal increases vulnerability of herring to oil spill effects compared to other finfish (Figure 5.1, Pearson et al. 1999, 2011). The eggs are subject to a number of stressors during incubation, and direct contact between the eggs and oil droplets is a condition likely to lead to poor hatching success and increased frequencies of abnormalities in the hatched larvae (Hay et al. 1995, Pearson et al. 1985, 1995). The brief near-shore occurrence of the yolk sac larvae and the near-shore rearing of Age 0 and Age 1 juvenile herring suggests that their vulnerability to oil is less than that of the eggs but more than that of the far-ranging adults. Pacific herring were of concern during three West Coast oil spills (Exxon Valdez, Nestuca, and Cosco Busan) and Baltic herring was of concern in the Tsesis spill in Baltic Sea. The Exxon Valdez Oil Spill (EVOS) was by far the most studied. Because of the prominence of the concerns for herring and the extent and length of the studies, this section will discuss the other three spills briefly and then examine the EVOS and herring in more detail.

July 2012 Page 5-23

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Figure 5.1 The Complex Life History of PWS Herring

Tsesis In October 1977, the tanker Tsesis grounded in the Baltic Sea and spilled 998 tonnes of Number 5 Fuel Oil and some Bunker C Fuel Oil (Linden et al. 1979, Teal and Howarth 1984). Clean-up collected all but 363 tons. Sonar surveys provided no evidence that herring schools avoided oiled areas (Linden et al. 1979). Frequency of Baltic herring spawning sites and hatching success of the eggs were lower in the oiled areas than in the reference areas, but conclusions could not be drawn because of inherent differences between oiled and reference areas (Linden et al. 1979, Nellbring et al. 1980).

Nestucca In December 1988, the barge Nestuca was struck by its tow and spilled about 782 tonnes of Bunker C Fuel Oil (Table 5.9). The oil stranded on beaches from Grays Harbor, Washington to , British Columbia. By the time oil entered Canadian waters, it had weathered substantially. Hay et al. (1995) studied the effects on herring eggs of Nestucca oil that had weathered naturally on eelgrass or rockweed for two to five months after the spill. Their

Page 5-24 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

laboratory experiments demonstrated decreased hatching rate and increased frequencies of abnormalities in hatched larvae for the eggs incubated on oiled substrate. Direct contact between the eggs and oil led to egg mortality.

Table 5.9 Summary Table for the Nestucca Oil Spill Oil Spill Name Nestucca Location Washington, USA Year 1988 Oil Type No. 6 Fuel Oil Specific Name Bunker C Fuel Volume (metric tonnes) 782 Platform Barge Environments Affected Marine Valued Ecological Component (VEC) Studied Fish Summary: On December 23, 1988 the fuel barge Nestucca collided with a tug and spilled approximately 734 metric tonnes of heavy fuel oil into the entrance to Grays Harbor off the coast of Washington. Most of the oil washed ashore near Ocean Shores, Washington but dispersed down to northern beaches and up to Vancouver Island, British Columbia totalling 1287 km2. There were no attempts for open water recovery of the spilled oil by U.S. or Canadian authorities. Extensive cleanup of affected beaches and wildlife including more than 13,000 birds were rehabilitated. In Barkley Sound in 1989, surveys did not find eggs deposited on oiled substrates but did reveal a change in the distribution of spawning from previous years. In 1990 and subsequent years, spawning distribution returned to its previous pattern (Hay et al. 1995, Hay et al. 2011). Estimated numbers of spawners in Barkley Sound was above average from the mid-1980s through the late 1990s (Hay et al. 2011).

Cosco Busan In November 2007, the container vessel Cosco Busan struck the Bay Bridge in San Francisco and spilled about 184 tonnes of Bunker Fuel Oil. Concerns quickly arose about the Pacific herring, which in spawns from December through March with a peak in February. In early 2008, Pacific herring eggs that had been artificially fertilized and then deployed sub-tidally in cages showed remarkably good hatching success (at or above 80%) in four oiled and two reference areas (NOAA/BML 2008). However, naturally fertilized eggs deposited on intertidal seaweed and rocks showed poor hatching success in four oiled areas compared to the one reference area. Hatching success of the intertidal eggs did not align with oiling levels, and chemical analysis of the intertidal eggs revealed a PAH signature indicative of urban runoff rather than Cosco Busan Oil. Despite the ambiguity of an effect attributable to the spill, concerns about effects on the 2007/08 year class that would recruit at Age 2 (in 2009/10) and low herring spawner abundance in the 2008/09, led the California Department of Fish and Game (CDFG) to close the 2009/10 herring season (CFDG 2009). The low herring abundance in 2008/09 could have been related to the

July 2012 Page 5-25

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

conditions then prevailing. Drought conditions enable high-salinity ocean water to intrude further than normal into the bay, and herring spawned much further up the bay in low- salinity water than in past years (CDFG 2009). Herring may skip spawning under adverse conditions (Engelhardt and Heino 2006; Kennedy et al. 2011). Spawn deposition surveys and test fishing in 2009/10 revealed increased herring abundance to a level approaching the long-term average, and the herring fishery in San Francisco Bay was opened again in 2010/11 (Figure 5.2; CDFG 2010). The 2010/11 spawning biomass was about 57,000 short tons “well above” the long-term average of about 49,000 short tons (CDFG 2011). This high biomass was supported primarily by Age 3 herring from the 2007/08 year class, which had been in the egg just after the spill. Whatever effects had occurred to the 2007/8 herring eggs and whatever the causes for the low 2008/09 spawner abundance, the herring population had returned to average pre-spill levels within two years and supported a commercial harvest within three years of the spill. San Francisco Bay herring abundance has been known to switch from low to high as ocean conditions change from El Nino to La Nina states (CDFG 2007; 2008), and the time course of the herring recovery appears to have been influenced by this transition in ocean conditions. The fishery closure also acted to accelerate rehabilitation of the stock from past harvest pressures on younger spawning herring and poor body conditions that preceded the spill (CDFG 2011).

Source: CDFG (2011) Figure 5.2 Spawning Biomass of Pacific Herring in San Francisco Bay

Page 5-26 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Exxon Valdez In late March 1989, the tanker Exxon Valdez grounded in Prince William Sound and released about 36,000 tonnes (10.9 million US gallons) of a light crude oil, Alaska North Slope Crude Oil (ANS) (NRC 2003). Pearson et al. (1995) found minor and localized effects on herring eggs but Brown et al. (1996) reported more widespread effects. One year later (1990), neither Pearson et al. (1995) nor Brown et al. (1996) found any effect on herring eggs. In 1990, adding Sitka Sound as a control area, completely independent of PWS, showed the profiles of hatching rates and other reproductive parameters essentially identical to those in oiled and reference areas of Prince William Sound (Figure 5.3; Pearson et al. 1995). In addition, there were high biomasses and harvests in the three years immediately following the spill (Pearson et al. 1999, 2011).

Figure 5.3 The Proportion of Total Eggs that were Developed, Hatched, Hatched with Viable Larvae, and Hatched with Larvae having Pericardial Edema The 1988 year class, which was Age 1 at the time of the spill and rearing in near-shore areas, recruited strongly as expected. Generally, a strong year-class recruits every four years in the PWS and other Gulf of Alaska herring stocks (Funk 1994, Williams and Quinn 2000a, 2000b). An estimated 2.1 billion Age 3 fish from 1988 year class recruited in 1991 compared to the 1.5 billion Age 3 age from 1984 year class recruited previously in 1987 (Funk 1994). The 1992 landings were 26,267 tonnes; the largest since the roe harvest started in early 1970s (Funk 1994). The 1993 herring season was predicted to have record high biomass and harvest (Funk 1993). What complicated perceptions about the recovery of PWS herring was a dramatic collapse of the biomass and fishery in the spring of 1993. Although a spawning biomass of 121,000 tonnes was predicted, only 27,000 tonnes were observed in spring 1993. Two hypotheses about the cause of the collapse emerged immediately: 1) that the EVOS was somehow responsible for the spill, and

July 2012 Page 5-27

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

2) that disease was the cause (Brown et al. 1996; Marty 1998; Pearson et al. 1999). The primary disease agent postulated at the time was Viral Hemorrhagic Septicemia Virus (VHSV). The virus is carried in the herring population and the disease is expressed when herring are confined or under other stress (Hershberger et al. 2006; Elston and Meyers 2009). Almost 20 years of research followed and at the end painted a more complicated picture of how other natural and human factors influence the population dynamics of PWS herring in ways that overwhelm any the oil spill effects and recovery (Pearson et al. 2011). The present consensus among investigators is that the EVOS did not cause the 1993 herring collapse in PWS. Most investigators dismissed direct causation because PAH levels in PWS water column had fallen to low background within a few years of the spill (Pearson et al. 1999) and because the lag between the spill and the collapse was too great (four years) (Rice and Carls 2007). In an experiment with other aims, Carls et al. (1998) exposed captive adult herring to the effluent of oiled gravel coated with weathered ANS and observed the onset of VHSV disease. Carls et al. (1998) provided a caution that the experiment did not demonstrate that the spill caused the decline. A series of three independent studies with five separate experiments (Kocan 1999; Kennedy 1999;, Sanders 2005) followed Carls et al. (1998) but failed to support the notion that exposure of low levels of oil could induce disease (Elston and Meyers 2010). Field data does support poor nutrition as the cause of the collapse (Pearson et al. 1999, 2011). First, the abundance of zooplankton prey for PWS herring was below average beginning in the mid-1980s (Figure 5.4, Cooney et al. 2002). Second, the gain in weight of adult herring over the winter began to decline in the mid1980s and between fall 1992 and spring 1993 the PWS herring lost weight (Figure 5.5). Two independent studies with Age Structured Assessment (ASA) models (Deriso et al. 2008; Hulson et al. 2008) provided critical examination of the contending hypotheses. The outcomes of these models were that the oil spill and harvesting effects were not the cause of the decline, rather poor nutrition and disease was. Because PWS herring have not returned to harvestable biomass levels, recent research has focused on the factors influencing the poor recovery including the oil spill (Pearson et al. 2011; Rice and Carls 2007). The poor recovery of PWS herring from the collapse is not due to the oil spill but to a combination of natural and human factors. Changes in ocean conditions in the GOA have influenced herring year class patterns in both PWS and Sitka. No strong year classes have emerged in GOA since 1993. The poor recovery cannot be due to overfishing since there was been no fishery in all but two years since the collapse. Similarly, with PAH levels at background since a few years after the spill and no overlap between spawning areas and the few places with oil remaining in beach sub-surfaces, there is no exposure to induce effects in the herring. Disease does appear to be present but not as a causal factor.

Page 5-28 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Source: redrawn by Pearson et al. (2011) from Cooney et al. (2001) Figure 5.4 The Settled Volume of Zooplankton in Prince William Sound

Source: drawn from ADFG data Legend: Squares Age 5, Crosses Age 7, Circles Age 9 Figure 5.5 Gain in Weight of Adult PWS Herring from Fall to Spring

July 2012 Page 5-29

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Although region-wide ocean conditions set the constraints on recruitment patterns in GOA herring stocks, additional factors must be influential on herring within PWS. Two factors appear to be influential. The abundance of overwintering humpback whales has increased and predation by humpback whales may be reducing adult PWS herring biomass (Rice 2008, 2009, Pearson et al. 2011). Interactions with juvenile pink salmon released from hatcheries (Figure 5.6) also affects the recruitment of PWS herring (Deriso et al. 2008) through predation on Age 0 herring and food competition with Age 1 herring (Pearson et al. 2011).

Source: ADFG data redrawn from Pearson et al. 2011 Figure 5.6 Spawning Biomass of PWS Herring and Number of Juvenile Pink Salmon Released by Hatcheries into PWS

Page 5-30 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

5.1.5.2 Anadromous Fish (e.g., salmon, steelhead) Because of their high commercial, ecological, and social value, concerns for salmon and other anadromous fish emerge quickly following an oil spill. Because the life history of pink salmon made them potentially vulnerable to oil spilled during the EVOS, this salmon species was well studied following the EOVS and is discussed below.

Pink Salmon and the EVOS The Exxon Valdez spill on March 24, 1989 was the largest oil spill in the at that time and there were widespread concerns that salmon resources in PWS would be seriously affected. The greatest concern involved the pink salmon, Oncorhynchus gorbuscha, the most abundant salmon species and the foundation of the commercial fisheries in Prince William Sound (PWS). Pink salmon have a consistent two-year life cycle, and thus are segregated as even- and odd-year runs (Heard, 1991). They spawn in freshwater, as do other salmon species, but the emergent fry go directly to the marine environment for their nursery period before embarking on a 14-month, open-ocean migration that ends with the return to spawn in their stream of origin. A large percentage of the pink salmon populations in PWS spawn in the intertidal reaches of streams entering the sound, and thus the incubation environment of the species was at risk of contamination from oil that washed into those stream reaches. The early marine life history of pink salmon also meant this species could be exposed to oil contaminated marine waters upon emergence as fry entering their nursery areas. However, there was great uncertainty about the actual level of risk that would be experienced by pink salmon. As the Exxon Valdez crude (EVC) reached landfall on southwestern shores of the Sound, the 1988 pink salmon brood year alevins were approaching the free-swimming fry stage and were exposed to oil in the incubation environment during that short interval of time prior to emergence from their streams in April and May, and subsequently in their marine nursery areas. The pink salmon of the following 1989 brood year experienced a different level of risk. They were exposed to residual oil that remained in the stream substrate during their entire incubation period from the fall of 1989 to the spring of 1990. Upon their emergence, oil in marine waters was nearly gone and the potential of risk on their early feeding life stage also diminished. Because of the uncertainty about the prospects of harm, scientists launched extensive studies to determine the effects of oil on incubating pink salmon eggs in PWS streams and on juveniles in the marine environment. Field studies on incubation success that included survival monitoring and in-stream bioassays revealed no elevated mortality of eggs or pre-emergent fry in the approximately 30 streams that were oiled out of approximately 1,300 streams entering the Sound, nor was there any post-emergent fry mortality observed in the receiving marine environment (Brannon et al. 1995; Moulton, 1996). This was not true in other studies that assessed the effects of EVC on pink salmon. A major effort was undertaken by Alaska Department of Fish and Game (ADFG) to monitor incubation success of pink salmon in selected oiled and non-oiled streams entering PWS. In the series of

July 2012 Page 5-31

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments stream surveys that were conducted over a nine-year period, ADFG reported higher egg mortality in oiled streams (Sharr et al. 1994; Bue et al. 1996, 1998), including some of the same streams in which non-agency scientists reported no spill effects (Brannon et al. 1995; Moulton, 1996). The Auke Bay Laboratory (ABL) of the National Marine Fisheries Service (NMFS), National Oceanic and Atmospheric Administration (NOAA), was also engaged in the assessment of spill effects on PWS pink salmon. NMFS researchers concentrated their effort on laboratory studies from which they developed an oil toxicity assessment. They concluded that very low concentrations of oil were responsible for the egg mortality observed by ADFG in PWS streams (Marty et al. 1997b; Heintz et al. 1999) and that exposure to EVC from the spill during incubation would have long-term effects (Heintz et al. 2000; Rice et al. 2001, 2007). In addition to the oiled stream studies the agencies undertook investigations on juvenile pink salmon in nearshore marine waters of PWS. Although their investigations did not reveal any mortality, they reported that in some cases juveniles associated with oiled marine areas of the sound were smaller than those in non-oiled areas. They interpreted smaller size to mean lower initial marine growth (Wertheimer and Celewycz 1996; Willette 1996). They felt smaller size would increase susceptibility to predation and result in lower marine survival based on other marine survival studies (Hargreave and LeBrasseur 1985; Heard 1991; Parker, 1971). ADFG took the egg mortality data and the growth data and applied them as variables in a model to predict what those effects would have on the number of returning adults (Geiger et al., 1996). The Geiger model predicted total effects of oil exposure on survival by using five stochastic steps: st = Femt × Fect × Dept × Segt × Smrt where st is the total survival, Femt is the fraction of the run that were females, Fect the mean fecundity, Dept the fraction of eggs actually deposited, Segt the egg survival, and Smrt the total marine survival. Using the Geiger model, the total mortality was estimated as a loss of about 2.3 million wild fish in the multiyear (1990–1994) return of 144 million adult pink salmon (Geiger et al. 1996). Although that number represented a relatively minor effect on the entire PWS pink salmon run, it constituted a major effect on those fish actually exposed to the EVOS. The estimated effect amounted to as much as 27% higher egg mortality in oiled streams as late as 1993. In 1989, the reduced size of juveniles exposed to oil in their marine nursery areas, combined with the egg losses, translated into an estimated 28% higher mortality of wild fish than what would have occurred in the absence of the spill. The Exxon Valdez Oil Spill Trustee Council (EVOSTC), a council of government agencies formed to assess the effect of the spill for recovery purposes, concluded from these data that the total mortality amounted to a reduction in returning adults of approximately 2% of the entire PWS pink salmon population and that the population did not recover from oil exposure until 2001, twelve years after the spill (EVOSTC 2002).

Page 5-32 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Oil Concentrations in PWS Streams The oil that was spilled from the Exxon Valdez was carried over 65 km in open water before it reached landfall, during which weathering exhausted much of the volatile compounds of the oil mass, reducing the monoaromatic components, and leaving the TPAH as the dominant component (Boehm et al. 1995). Because of tidal dilution and freshwater outflow of the affected streams flushing the intertidal substrate during receding tides, the actual deposit of oil in the gravel of salmon streams was very low. Mean sediment-TPAH concentrations in oiled streams, when oil was at its highest concentration in 1989, ranged from 0.5 ppb–267 ppb (Brannon et al. 1995), or only small fractions of the lethal sediment-TPAH threshold of 3,800 ppb–8,300 ppb determined by laboratory bioassays on chronic exposure of pink salmon eggs (Heintz et al. 1999; Brannon et al. 2006a). Mean sediment–TPAH levels in non-oiled reference streams was less than 65 ppb. In the subsequent years, the 1990 oiled streams still showed that low concentrations of residual oil were present in sediments with mean TPAH ranging between 1 ppb and 413 ppb, with one exception (2,818 ppb from one sample), and in the following year means dropped to between 1 ppb and 236 ppb. These data show that the contamination of the intertidal stream reaches during the time that oil was present in the stream substrate was well below those concentrations imposing risk of pink salmon incubation mortality in laboratory bioassays (Brannon et al. 2006a; Heintz et al. 1999; Marty et al. 1997b).

Tissue-TPAH Analysis of Embryos Sampled from Oiled Stream Mean tissue-TPAH concentrations in alevins sampled from oiled streams in 1990 and 1991 were 62.8 ppb and 94.5 ppb dry wt, respectively, and slightly higher than the mean tissue–TPAH of alevins in non-oiled streams of approximately 39 ppb over the same period (Brannon et al. 1995). These concentrations are far below lethal tissue–TPAH threshold concentrations greater than 7,800 ppb shown in long-term laboratory bioassays with pink salmon embryos exposed to naturally weathered EVC (Brannon et al. 2006a). The tissue data of alevins sampled from oil steams were equivalent to the laboratory controls (approximately 90 ppb) that experienced no oil exposure in respective bioassays of Heintz et al. (1999) and Brannon et al. (2006a).

No Pre-Emergent Fry Mortality Observed in Field Studies Compelling evidence against oil as a threat in the incubation environment was the lack of any difference in survival of the pre-emergent fry stage between oiled and non-oiled streams in the spring following elevated egg mortality in the fall (Bue et al. 1996). That was also shown by other investigators monitoring survival in oiled and reference streams, including some of the same streams reported to have high egg mortality, with no statistically significant differences at the emerging fry stage (Brannon et al. 1995; Sharr et al. 1994). This result was counterintuitive because alevins should be especially sensitive to acute oil toxicity (Moles and Rice, 1983). If oil was responsible for egg mortality, not only should the effects of toxicity have continued, but the mortality should have translated into much lower numbers of fry. This suggested that the increased mortality in the fall surveys was confined only to those eggs extracted from the redds, and not among those that continued incubation to the spring fry stage.

July 2012 Page 5-33

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Absence of Anomalies among Naturally Incubating Alevins Alevin samples removed from the incubation substrate in the 1990 by Brannon et al. (1995) were examined for developmental anomalies. No difference (p > 0.05) in the proportion of normal alevins in oiled (0.989) versus reference (0.992) streams was seen. The absence of anomalies from field samples contrasted with what was observed in laboratory studies (Brannon et al. 2006a; Marty et al. 1997a), where deformities and blue sac disease (ascites) were observed in alevins exposed to laboratory TPAH doses. The presence of induced deformities and blue sac disease in the laboratory and their absence in the field over the same length of time indicate that laboratory test doses were not representative of that which existed in the oiled streams of PWS.

Retarded Growth Not Shown in PWS Emerging Fry In laboratory studies at ABL, growth of embryos was affected by exposure to high oil/gravel mixtures (Heintz et al. 1999; Marty et al. 1997b) with a slightly slower rate of development and greater amounts of yolk remaining in the abdominal cavity at emergence compared to controls examined at the same time. This resulted in a slightly greater weight per unit of length, and thus a higher developmental index (kD = 10 × [wt in g]3/l in mm, Bams 1970). Marty et al. (1997b) suggested this may be a practical assessment of oil effects. Therefore, in 1989 and 1990 when emerging fry from oiled streams were statistically no different in length or weight than in non- oiled streams, and thus no different in their developmental index, this was another demonstration that an oil effect in the field was unlikely. In 1990, fry from reference streams showed a mean kD of 1.76 and fry from oiled streams were at a mean of 1.79, indicating unaltered emergence timing (Brannon et al. 1995). Premature emergence would have been evident by a higher kD index (greater than 1.9) and that wasn’t the case.

Concurrent Incubation Studies in 1989 In the fall of 1989, Moulton (1996) conducted a field study on egg survival at the same time and in some of the same oil contaminated streams reported by Bue et al. (1996, 1998) to have higher egg mortality. Moulton artificially spawned adults from each of nine streams, placed the fertilized eggs in 24 perforated incubation boxes per stream, and buried the boxes in the intertidal stream substrate. The incubation boxes isolated effects that would be associated with oil, such as leachate emanating from oil deposits entering redds (Murphy et al.1999; Rice et al. 2001), from the lethal effects of over-spawning (Collins et al. 2000) and any deleterious effects of sampling that were unrelated to oil (Brannon et al. 2001). Upon approaching the hatching stage, the eggs were removed from the incubation boxes and embryo survival was assessed. Moulton (1996) found that survival was high, averaging 96.5% in oiled streams and 95.1% in non-oiled reference streams (p = 0.870). This was evidence that demonstrated embryo survival was unaffected by the oil concentrations that existed in the intertidal reaches of those streams at the height of contamination of beaches and streams along the spill-path.

Page 5-34 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Summary for Anadromous Fish Although deposits of oil on the shorelines were hypothesized to produce leachate resulting in interstitial toxic water entering the redds and causing embryo deaths, actual field data show that was not the case. As confirmed in the field studies, HPAH compounds alleged to have been the toxic components of weathered oil are actually increasingly less available to incubating embryos as oil weathers and thus not the source of toxicity. Exposure to oil in PWS streams was not related to any damage that could be identified in pink salmon during or after the incubation experience, and there is no evidence that adults exposed to oil as embryos in oiled streams experienced lower return rates. The long-term effects of the Exxon Valdez oil spill on the PWS pink salmon survival is best demonstrated by the success of adult returns. The general good health of the PWS population immediately following the spill is shown by the record returns of pink salmon to the sound in 1990 and 1991 (Figure 5.7), which were the returns from embryos exposed to the oil in 1989 and 1990 (Brannon et al. 2006b), and their successful productivity since then by nine additional record years. Nearly 3.5 times the number of 1988 adults returned in 1990, although 6% fewer fry were released from the PWS hatcheries that produced that large run. A stock-specific run reconstruction model for PWS pink salmon by Templin et al. (1996) also concluded there was no reduction in adult returns from the spill. These data present a compelling case that pink salmon were not measurably damaged by the Exxon Valdez oil spill. There is no evidence supporting the projected losses to the PWS pink salmon that were anticipated at the time of the spill event or estimated by the Geiger model.

Figure 5.7 Adult Pink Salmon Returns to PWS from 1960–2007

July 2012 Page 5-35

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

The large hatchery programs in the Sound are major contributors to the success in cultivating the large runs of PWS pink salmon. Natural variability will occur in numbers of adults returning as a reflection of PWS productivity, such as the low returns in 1992 and 1993 caused by low temperatures affecting growth and reduced plankton productivity (Cooney and Willette 1997; Willette 1992). But, the success of the hatchery runs is also evidence that oil passing through their marine nursery areas and migratory pathways in the spring of 1989 did not inhibit the record returns of those fish in 1990. It also shows the demonstrated success of the natural returns to oiled streams in the twelve years of study thereafter (Brannon et al. 2006b; Maki et al. 1995). Recovery efforts for pink salmon were not applied by any state or federal agency nor were they necessary because the evidence shows they were not damaged in any measurable way.

5.1.6 Marine Birds Marine birds can be affected by oil by way of several pathways (direct contact, ingestion, inhalation of volatiles, absorption through skin). Direct contact with oil can destroy the waterproofing abilities of feathers and may cause death by hypothermia. Additionally, preening of soiled feathers, in an attempt to remove oil, can lead to ingestion and asphyxiation or adverse health effects including mortality. Birds can also transfer oil from their feathers to the surface of their eggs during incubation and, depending on the type of oil, embryos may fail to develop (NRC 2003). Adverse effects from oil spills on the distribution, abundance, and availability of prey can also lead to indirect effects on reproduction and survival (e.g., Piatt and Anderson 1996; Peterson 2001; Golet et al. 2002; NRC 2003). Following an oil spill, the extent of marine bird mortality depends on a range of factors, including the location and timing of the oil spill (e.g., seasonal aggregations of birds occur for breeding and migration), the size and density of the local marine bird population, the foraging behaviour of the species present, and the quantity and persistence of the spilled oil (Piatt et al. 1990). The most vulnerable taxa are thought to be loons, grebes, sea ducks, and alcids (e.g., pigeon guillemots) because they spend the majority of their time swimming on the sea surface and often aggregate in dense flocks (Piatt et al. 1990). Following the 1989 Exxon Valdez oil spill (EVOS) in Prince William Sound (PWS), Alaska, an estimated 30,000 to 52,330 carcasses from 90 bird species were reportedly retrieved from the spill area between early April and late September 1989 (Piatt and Ford 1996; Wiens 1996; Piatt et al. 1990). It is not known to what extent these numbers represent acute effects directly as a result of the EVOS. Survey data indicated that a proportion of the carcasses recovered after August 1989 had become oiled after mortality from other undetermined causes (Piatt and Ford 1996, Piatt et al. 1990). Statistical models completed by Ford et al. (1991) assessed mortality at 375,000 birds based on estimated parameter values and the number of recovered carcasses. Other scientists estimated mortality from alternative models (Piatt and Ford 1996, Heinemann 1993, Piatt et al. 1990) that yielded 100,000 to 690,000 birds. However, due to a lack of baseline data on avian populations in PWS at the time of the spill, the actual extent of mortality is not known.

Page 5-36 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Of the 30 000 dead birds that were identified, 74% were murres, 7% were other alcid species, and 5.3% were sea ducks. Eighty-eight percent of birds were retrieved outside PWS (Piatt et al. 1990). Decreases in local populations of loons, grebes, cormorants, and sea ducks ranged from 44-84% based on pre- and post-oil contact aerial surveys, while local gull and eagle populations increased from 87-240% (Piatt et al. 1990), likely due to the increased opportunities for scavenging dead birds and other animals. Research has resulted in a general consensus among scientists that bird species responded differently to the EVOS due to specific life history characteristics that dictated potential exposure and vulnerability (Esler et al. 2011; Esler et al. 2010a; Day et al. 2003; Day et al. 1997). Following the EVOS, Murphy et al. (1997) examined changes in overall marine bird abundance of 12 taxa among oiled and unoiled sites in PWS from 1984-1985 (pre-spill) to 1989, 1990, and 1991 (post-spill). No change was observed for seven taxa, including common mergansers, black oystercatchers, Bonaparte's gulls, mew gulls, black-legged kittiwakes, and murrelets. Three species (red-necked grebes, pelagic cormorants, and pigeon guillemots) exhibited decreased overall abundance in all three post-spill years (Murphy et al. 1997). However, as reported previously (Piatt et al. 1990), bald eagles and glaucous-winged gulls exhibited increased abundance and Murphy et al. (1997) concluded that by 1991, signs of recovery (based on abundance and distribution) were evident for all affected taxa. In addition to reductions in population sizes, research showed direct and pronounced negative effects on habitat use by many marine bird species, with the strongest effects seen in species associated with the shoreline, or which were year-round residents of PWS or the Gulf of Alaska (Day et al. 1997a, 1997b; Wiens et al. 2001; Murphy et al. 1997). These findings were supported by guild analysis, which indicated that the most consistent effects from oiling were on species that fed on or close to shore, bred on the beach, or were winter or year-round residents (Wiens et al. 1996). Later studies used multiple years of survey data, in addition to information on habitat features, to examine relationships between habitat type and degree of oiling (Wiens et al. 2004). One model treated the degree of oiling as a quantitative variable, with habitat measures included as covariates. The second model treated oiling as a categorical variable to examine abundance trends with 1984 data as a baseline. Nine species that were found to have suffered adverse effects to habitat occupancy as a consequence of oil spills (estimated date of cessation of spill effects on habitat use is indicated in brackets): • Common mergansers (1996) • Spotted sandpipers (1990) • Belted kingfishers (1996) • Steller's jays (1991) • Northwestern crows (1990) • Mew gulls (1991) • Bald eagles (1990)

July 2012 Page 5-37

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

• Pelagic cormorants (1991) • Pigeon guillemots (1991, although lower abundance in oiled bays in 1996 and 2001 could indicate a delayed, inconsistent spill effect) (Wiens et al. 2004) Some of the dates, referred to above, are later than estimates provided by Day et al. (1997b), who concluded that by 1991, there were no obvious ecological impediments to the recovery of habitat use by species that had not yet recovered from the effects of the spill. Conversely, differences in habitat features, rather than the degree of oiling, was found to be the cause of variation in abundance between oiled and unoiled bays for 11 species (great blue herons, Barrow's goldeneyes, common ravens, red-necked phalaropes, glaucous-winged gulls, marbled murrelets, black-billed magpies, wandering tattlers, common murres, tufted puffins, and common goldeneyes) (Wiens et al. 2004). Despite the apparent return of habitat use, Lance et al. (2001) reported that as of 1998, population trends within oiled areas showed no evidence of recovery for grebes, cormorants, black oystercatchers, mew gulls, glaucous-winged gulls, terns, murres, pigeon guillemots, or murrelets. Further, based on population trends, they reported evidence of continued effects in scoters, mergansers, goldeneyes, and black legged kittiwakes. Lance et al. (2001) reported that loons, harlequin ducks, buffleheads, and northwestern crows showed evidence of recovery. Irons et al. (2000) compared marine bird density in PWS pre-spill (1984-85) and post-spill (1989-1991, 1993, 1996, and 1998) for 14 taxa and found lower than expected densities through 1998 for cormorants, goldeneyes, mergansers, pigeon guillemots, and murres. Whereas, black oystercatchers and harlequin ducks demonstrated adverse effects in 1990 and 1991 (Irons et al. 2000). Glaucous-winged gulls, murrelets, and terns showed relative increases in most post-spill years. Thus, a combination of potential lingering spill effects and natural variability (e.g., forage fish abundance) may have been acting to delay recovery of these species (Lance et al. 2001; Irons et al. 2000). While these marine bird surveys provided valuable information on habitat use and population trends in numerous species, a more in-depth examination of species-specific responses revealed that a complicated interplay of factors (both anthropogenic and natural) may delay population recovery following an event, such as the EVOS. Relevant case studies are presented in the following sections for harlequin ducks and pigeon guillemots because there is a long and thorough body of peer-reviewed literature for both species.

5.1.6.1 Pigeon Guillemots Pigeon guillemots are semi-colonial sea birds and members of the auk family (Alcidae). They are endemic to the North Pacific Ocean, ranging from north of the Bering Strait south to Santa Barbara Island, California (Ewins 1993). Their preferred habitat occurs along rocky coastlines, which provide suitable nesting sites and shallow water for foraging (Ewins 1993). Pigeon guillemots are diving foragers, consuming primarily small fish and invertebrates. Eggs are laid from early May to mid-June in California and British Columbia, (at least a week or two later in Alaska), and chicks stay in the nest for approximately 30-40 days (Ewins 1993).

Page 5-38 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Since 1972, pigeon guillemot populations in the Bering Sea and the Gulf of Alaska have experienced substantial declines. Specifically, in PWS, the population declined by approximately 72% between 1972 and 1989-1993 (Agler et al. 1999). The Exxon Valdez oil spill (EVOS) is believed to have resulted in mortality of approximately 500 – 1500 birds (Litzow et al. 2002), which represented 5-14% of the overall decline (Piatt et al. 1990). The majority of the population decline, however, has been attributed to the Pacific Decadal Oscillation (PDO), which, in 1977, resulted in an over 90% loss of lipid-rich capelin populations and an increase of 250% in populations of lipid-poor demersal fish (largely gadids and pleuronectids) (Litzow et al. 2002). A subsequent oceanic regime shift in the late 1980s did not return the Gulf of Alaska food web to the pre-1977 conditions (Litzow et al. 2002). The decline of pigeon guillemot populations in PWS was concurrent with a decline in the proportion of lipid-rich Pacific sand lance (Ammodytes hexapterus) in the diets of chicks and replacement with a variety of low-lipid demersal fishes (Golet et al. 2002; Peterson 2001). Consumption of high-lipid forage fishes is positively correlated with chick growth rates, nestling survival, and productivity (Golet et al. 2000), and the increased proportion of low-lipid fishes in the post-spill diet was associated with decreased survival of nestlings at oiled sites compared with unoiled sites (Golet et al. 2002). It has been hypothesized by Golet et al. (2002) that declines in sand lance abundance, which occurred before the EVOS, may be constraining the ability of pigeon guillemot populations to recover from the EVOS by limiting the ability of the population to replace individuals lost to oil pollution. In the 10 years following the EVOS, pigeon guillemots inhabiting oiled areas of PWS exhibited elevated levels of cytochrome P450 1A, a biomarker associated with oil exposure, compared to birds from unoiled areas (Trust et al. 2000). In addition, a strong positive relationship between cytochrome P450 1A (CYP1A) levels and levels of aspartate aminotransferase (AST) was indicative of organ damage (Golet et al. 2002). However, the interpretation of these results may be considered debatable since AST was significantly elevated in only one of the two study years. Only adults, not chicks, exhibited elevated CYP1A levels, which was attributed to divergent diets (Golet et al. 2002). Chicks consume only fish, which are effective hydrocarbon metabolizers and would thus not be expected to bear a hydrocarbon burden. Conversely, adult guillemots consume fish and also benthic invertebrates, which generally cannot metabolize hydrocarbons effectively (Meador 2003), and may act as an important dietary source of hydrocarbons. Pigeon guillemot populations in PWS were labeled ‘not recovering’ in 2010, over 20 years after the spill (EVOSTC 2010). However, population declines have been observed in both oiled and unoiled areas of PWS, and the decreased availability of high-lipid forage fish is assumed to be an important contributing factor (Golet et al. 2002).

July 2012 Page 5-39

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

5.1.6.2 Harlequin Ducks Following their breeding season, which occurs inland in fast-flowing rivers and streams, Harlequin ducks occupy shallow intertidal zones along rocky coastlines of North America (Robertson and Goudie 1999). As a species, they exhibit high site fidelity and relatively low annual productivity, which is compensated for by relatively high adult survival and a long reproductive life span (Iverson and Esler 2010; Esler et al. 2002). As a consequence of their small body size, harlequin ducks are thought to exist near an energetic threshold in winter (Esler et al. 2002). In particular, PWS is the northern extent of their habitat range, and where it is thought that the additive effects of harsh winters and the high metabolic cost of dive-foraging reduce their resilience to natural and anthropogenic environmental perturbations (Iverson and Esler 2010). Harlequin ducks are also vulnerable to oil pollution as a consequence of their diet. A nearshore predator, they forage in the intertidal zone, consuming benthic invertebrates including amphipods, snails, and crabs (Robertson and Goudie 1999). These organisms are inefficient metabolizers of hydrocarbons (Meador 2003), and as such can act as a hydrocarbon source. Further, harlequin ducks come into contact with potentially contaminated sediments when foraging for sediment-associated benthic invertebrates (Iverson and Esler 2010). Immediately following the EVOS spill, 212 harlequin duck carcasses were recovered (147 of these were in PWS). Using a 15% recovery rate (Piatt and Ford 1996), estimated that total losses were 1 413 birds, with 980 of those in PWS. This value represented approximately 7% of the PWS Harlequin duck population (Esler et al. 2002). Since the EVOS, several hypotheses to explain slow or absent recovery of the PWS harlequin duck population have been investigated. These include differences in habitat, prey availability, and body condition, biomarker studies to examine the possibility of ongoing exposure to residual oil, and ongoing demographic effects. There is considerable disagreement among scientists on the relative importance of habitat versus degree of oiling in determining harlequin duck exposure histories and recovery probabilities. For example, Esler et al. (2000a) examined population densities in PWS from 1995-1998 in relation to habitat characteristics including substrate, exposure, intertidal slope, prey biomass, and degree of oiling by the EVOS, and found that harlequin duck densities in winter were inversely proportional to oiling history even after these other variables had been accounted for. In addition, an absence of differences in prey density and total biomass (summer 1997) and harlequin duck body mass (late summer/early fall 1995-97; winter 1997-98) between oiled and unoiled sites indicated that these factors were not contributing to a lack of population recovery (Esler et al. 2000a). Conversely, using statistical modeling that incorporated survey data from PWS, Wiens et al. (2004) found habitat variables, rather than the degree of oiling, to be a better predictor of harlequin duck abundance among study areas. They concluded that between 1991 and 1996, there was no evidence of an effect of oil on habitat occupancy by Harlequin ducks (Wiens et al. 2004).

Page 5-40 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

In the ten years following the EVOS, Harlequin ducks inhabiting oiled areas exhibited elevated of CYP1A relative to those from unoiled areas (Trust et al. 2000). The authors also examined PCB exposure, which can induce CYP1A activity, by measuring PCB concentrations in plasma samples. However, PCB levels could not account for the spatial differences observed in CYP1A induction (Trust et al. 2000). Follow up work spanning the period from 2005 to 2009, revealed that Harlequin ducks from oiled areas continued to exhibit higher levels of CYP1A activity than those from unoiled areas (Esler et al. 2010), suggesting that up to 20 years after the initial spill, Harlequin ducks in some parts of PWS continued to be exposed to residual oil. Whether the EVOS is a source of chronic contamination remains uncertain because the trace oil detections recorded during studies may potentially be from other chronic sources, such as routine vessel discharge, or various sites where there are natural coastal oil seeps (Irons et al. 2000). Demographic effects also appeared to be constraining the recovery of the Harlequin duck population in PWS. Between 1995 and 1998, adult females from oiled areas experienced lower winter survival rates than those observed in females from unoiled areas (Esler et al. 2000). However, follow up work in the winters of 2000-2001 and 2002-2003 found that these effects had largely abated (Esler and Iverson 2010b). These findings were subsequently used in matrix population modeling in an attempt to project a recovery timeline for the PWS Harlequin duck population (Iverson and Esler 2010). The model indicated that under a ‘most likely’ combination of variables, recovery would require 24 years, with a range of 16 to 32 years under best- and worst-case scenarios, respectively (Iverson and Esler 2010).

5.1.6.3 Other Marine Bird Species and the EVOS Although Common Loon populations were reported to have declined immediately following the oil spill event (Irons et al. 2000), long-term surveys conducted from 1989 to 2007 found increasing winter population trends were returning to pre-spill levels. Loon populations appeared to begin recovering in 1991 and are now at, or have exceeded, historical numbers (EVOSTC 2010, Day et al. 1997b). Productivity of Bald Eagles (i.e., numbers of young successfully fledged) was reduced in oiled areas of Prince William Sound in 1989; 30% of occupied nests produced young. Aerial surveys confirmed productivity had returned to normal in 1990 and 1991 (Bowman 1999, Day et al. 1997b) and by 1995, reproductive success and/or recruitment was estimated to have returned the population to or exceeding its pre-spill numbers (EVOSTC 2010). Although the timing and success of breeding at murre nesting colonies within the Gulf of Alaska was reportedly disrupted, reproductive success was recorded at pre-spill population abundances from 1993 to 1997 suggesting that recovery took approximately four to seven years (EVOSTC 2010). Day et al. (1997b) determined that Glaucous-winged Gull initially showed adverse effects with no evidence of recovery. Subsequent publications indicated substantial population increases in Prince William Sound, which may partially be attributed to the indirect effects of vessel clean-up activities which tend to provide sustenance to foraging gulls (Irons et al. 2000, Murphy et al.

July 2012 Page 5-41

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

1997). Overall, the populations of Glaucous-winged Gull in Prince William Sound are considered now to have re-established to pre-spill levels (Irons et al. 2000). Recent studies suggest that breeding colonies of Black-legged Kittiwake in the Prince William Sound have since recovered and increased 1.6% since the 1970’s (USFWS 2006). As such, it is expected that Black-legged Kittiwake has recovered to pre-EVOS levels (USFWS 2006). The population of Barrow’s Goldeneye in the Gulf of Alaska was considered by some scientists to have recovered from the EVOS by 1996 to 1998 (Wiens et al. 2001). Others consider the population to be increasing and still recovering from the effects of the oil spill (EVOSTC 2010). Black Oystercatcher population numbers are considered by some scientists to have recovered (Murphy and Mabee 2000; Agler et al. 1999), while other scientists have reported population trend estimates that indicate numbers have not returned to pre-spill levels (EVOSTC 2010). The rate of population declines of Kittlitz’s Murrelet in PWS were estimated at 18% in 1972 but increased to 31% after the EVOS in 1989. The recovery status of this species is uncertain and the factors contributing to the continued decline of the population remain unresolved (EVOSTC 2010).

5.1.7 Marine Mammals

5.1.7.1 Cetaceans A leaking oil well in Santa Barbara Channel off southern California in 1969 was the incident that first raised concerns about the possible effects of spilled oil on cetaceans (Geraci 1990). At the time of that spill, Gray Whales were just starting to migrate north from the calving lagoons of Baja California to feeding grounds, primarily in the Bering and Chukchi seas. Some of these whales were observed to swim through the slick, and over the course of the spill, six dead whales were found, but only two were fresh enough for necropsy. Oil was not found on either one, and it was concluded that they had died of natural causes, although one may have been harpooned (Brownell 1971). Some of the news accounts at the time reported dead dolphins that were heavily oiled, but the biologist who examined several of the dead dolphins found no evidence of oiling (Brownell 1971). Geraci (1990) comprehensively reviewed all of the information available on cetaceans and their responses to oil spills. These spills included that of the Argo Merchant, Amoco Cadiz, Ixtoc-1 (well ), Regal Sword, Hellenic Carrier, and Alvenas, among others. In several of these spills, whales and/or dolphins were observed swimming in oiled waters. After evaluating the evidence, Geraci (1990) concluded that “… there is no gripping evidence that oil contamination has been responsible for the death of a cetacean.” Since Geraci’s review oil spills have occurred, including the Exxon Valdez, Prestige, and Jessica. With the exception of the Exxon Valdez (discussed below) Geraci’s overall conclusion still stands. That is, despite observations of cetaceans in a number of spills, the effects, if any, have not been apparent.

Page 5-42 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Exxon Valdez and Cetaceans Following the EVOS, biologists mounted several efforts in attempts to determine whether cetaceans had been affected. First, by the day following the spill, biologists started aerial surveys of the spill area, and vessel surveys began on day 4 (Zimmerman et al. 1994). Second, ongoing research on killer whales was continued. As in British Columbia, many of the killer whales that used Prince William Sound, Alaska, were known as individuals, based on pigmentation patterns and other marks, and had been followed for several years. Consequently, age (often approximate), diet, sex composition, contaminant burdens, membership in social groups, and interactions with human activities, particularly fishing, were at least partially understood. The Alaska studies began in about 1984, although some information had been collected earlier (Matkin et al. 1994, 2008; Leatherwood et al. 1990).Third, vessel surveys dedicated to documenting oil effects on cetaceans were conducted in the oil-affected area from 1-9 April (Harvey and Dahlheim 1994). These were general searches for cetaceans that might have been in oiled waters. Fourth, humpback whale studies, based on photo-ID, which had begun in about 1980, and had been continued in 1988, were then resumed for 1989 and 1990 to look for possible effects of the EVOS. Fifth, aerial shoreline surveys were conducted in PWS and northern Gulf of Alaska in an effort to locate fresh carcasses of whales and dolphins that may have been killed by oil exposure; when possible, the carcasses were visited for closer examination and the collection of samples (Loughlin 1994). Except for the results of studies of killer whales (see below), the data gathered fell within apparently normal bounds.

EVOS and Killer Whales The Exxon Valdez oil spill is, to date, the best-studied major tanker , and the findings of studies related to killer whales suggested that there had been adverse effects (Matkin 1994, 2008). Just after midnight on 24 March 1989, the tanker Exxon Valdez ran aground on Bligh Reef in Prince William Sound, Alaska (Morris and Loughlin 1994). Over the next five hours, oil flowed from the tanker, while at the same time liberating VOCs (volatile organic compounds) (Galt et al. 1991; Hanna and Drivas 1993). Under the relatively quiet conditions that prevailed, the slick that formed spread under the influence primarily of gravity (NRC 2005). Under such conditions, oil will continue to spread until it reaches a thickness of about 0.1 mm, with considerable loss of the lighter organic compounds. Late on day 3 of the EVOS, a storm developed, which carried the oil to the southwest, where it encountered a number of islands (Galt et al. 1991). Under the influence of high winds, evaporation of the remaining VOCs increased, as did dispersion into the water column, and the formation of an oil-in-water emulsion (mousse) was promoted. Much the mousse was driven onto the shorelines of the islands of southwestern PWS where it accumulated, and under the influence of both wind and tides, the material worked its way into the nearshore and beach material. Most of the open-water areas that were affected by the spill and where the killer whales may have been present, had only a very thin sheen of oil whose thickness would have been measured in microns (Galt et al. 1991).

July 2012 Page 5-43

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

During the first week of the EVOS, aerial and boat surveys were conducted for marine wildlife, but no killer whales were sighted. However, a photographer from the Los Angeles Times documented the presence of killer whales from the AT1 group near the stricken ship on day 2. It was not until day 7 that whale biologists found whales from several other groups in waters with an oil sheen, but this was at a location about 100 km southwest of the spill site. By this time, the oil had weathered considerably and had lost nearly all of its VOCs. Along the shoreline, under the influence of wind and tides, oil and mousse accumulated, affecting the nearshore environment. Although killer whales are still technically treated as though they were a single species, genetic results now point to a different conclusion. Two forms of killer whale inhabit Prince William Sound: the resident, fish-eating form, and the transient, marine-mammal eating form (Matkin et al. 1999, 2008). Recent genetic analyses have shown that, despite their outwardly similar appearance, these forms are full species, not having interbred for perhaps 700,000 years (Morin et al. 2010). It is notable that these forms are ecologically very different and distinct, with the residents feeding on primarily on salmon, while the transients feed on marine mammals, particularly harbour seals and Dall’s porpoises, which are at the next higher level in the food chain than are salmon. Consequently, the transients are exposed to much higher levels of POPs (persistent organic pollutants), which have been bio-magnified through the food chain and have been implicated in reproductive and immune function failure (Ross and Birnbaum 2003, Ylitalo et al. 2001). In addition to suffering large mortalities that were discovered after the EVOS, AB pod differs from other resident killer whale pods in PWS in that it was also implicated in depredating longline fisheries for sablefish and halibut (Fraker 2012; Matkin et al. 2008). As a consequence of this interaction, fishermen shot a number of the whales with high-power rifles. Some of the whales died in 1985 and 1986, presumably as a consequence of the bullet wounds, and some of the later mortalities, discovered following the EVOS may have been caused by bullet wounds as well (Fraker 2012). Of the seven AB whales that were discovered dead in 1989, four had documented bullet wounds. Because cetacean skin heals quickly or perhaps because some bullet wounds may not have been located where they could be easily photo-documented, it is possible that additional whales had also been wounded. In fact, the bullet wounds on some of the whales photo-documented to have been wounded became undetectable within a few months (Fraker 2012). The level of detail in the record for certain killer whale groups in PWS is impressive. For the resident whales, the composition of the individuals comprising seven pods are fully documented in most years. Transient whales are generally more variable in their occurrence, with gaps in the record of several years being common, so the record is less complete for them. An exception is the AT1 group, in which most, but not all, members are observed in most years. As mentioned above, however, there are long gaps in the record, and this places limits on what can be reliably be concluded about the timing and causes of death.

Page 5-44 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Mortalities amongst Killer Whales All researchers investigating the killer whale deaths that were discovered following the spill agreed on the basic facts: • AB pod (fish-eating, resident): seven deaths were discovered in 1989, six in 1990 • AT1 group (marine-mammal eating, transient): nine deaths were discovered in 1990 The word “discovered” is used here because there are gaps of several months in the sightings records; few records are made during “winter” (September – May) and even during summer, the whales are not usually accounted for most of the time. For example, the whales that were discovered missing 7 days after the spill had not been seen for more than 6 months, so whether they died before or after the spill, all at the same time, or from the same or multiple causes are all unknown (Fraker 2012). As another example, there is a gap of nearly four months (31 March – 27 July) in the 1989 record for AB pod (Fraker 2012)); where the whales were for most of the 5 months after the spill is unknown. In the Pacific Northwest, studies of killer whales rely on photo-documentation, based on individually distinctive pigment patterns and scars. Documenting dead killer whales is much more difficult because fewer than 10 % strand on beaches where they can be examined, and it is rarer still to find the carcasses when they are fresh enough for a full necropsy. Consequently, the 7 AB-pod deaths uncovered in 1989 were documented on the basis of missing animals; no bodies were found and neither the timing nor cause-of-death is known. The remaining deaths (both AB pod and AT1 group) did occur after the EVOS (sometime between September 1989 and June 1990), but exactly when, where, and why are unknown. The one AT1 carcass that was found was too badly decomposed to yield useful information on cause of death.

Fishery Interactions with Killer Whales In the mid-1980s, killer whales of AB pod were involved in depredating longline fisheries that were targeting sablefish and halibut. During this period, the whales consumed an estimated one- quarter of the catch, which constituted a major economic loss to the fishermen (Dahlheim 1988). This led the fishermen to shoot the whales, particularly in 1984 and 1985, and six whales died during this time, although not all of them were documented to have been wounded. An additional nine whales were wounded, but survived, at least in the near term (Matkin et al. 1988). There were additional deaths in 1986 and later, some of which may have been related to the gunshot wounds. Although the regulations changed in 1987 to prevent shooting of whales by fishermen, some shooting has continued (Fraker 2012).

Exposure of Killer Whales to Oil If oil was responsible for some or all of the deaths discovered following the EVOS, a plausible route(s) of exposure is required. That is, there must be at least some evidence that the whales were exposed to potentially harmful amounts of petroleum compounds. But it was not until the end of the first week that biologists found killer whales in oil-contaminated waters, and this was after the storm and in an area about 100 km from the spill site. A photographer from the Los

July 2012 Page 5-45

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Angeles Times did, however, document the presence of killer whales near the Exxon Valdez on day 2 of the spill, and some of these were determined to be missing one year later. Matkin et al. (2008) implied that VOCs were likely to have been involved, and certainly, exposure to high concentrations of petroleum vapours can be fatal either directly or indirectly from narcosis, which would make the animals vulnerable to drowning (Geraci and St Aubin 1990). The concentration of VOCs in the EVOS was highest during the first six hours of the event, and to have been exposed to the greatest concentrations, the whales would have to have been in the middle of the spill at that time (when it was dark and no one would have been able to see them). But the maximum total VOC concentrations estimated for the EVOS was less than 30 ppm (Fraker 2012, Hanna and Drivas 1993), which is an order of magnitude below the levels (300 ppm) that Geraci and St Aubin (1987) estimated to be necessary to cause neurologic effects that might have led to drowning. Lower concentrations would have resulted in mild and reversible irritation to sensitive tissues. Part of AB pod was observed in waters with an oil sheen a week after the spill, which prompted Matkin et al. (2008) to state that the whales had been “… unequivocally documented … [in] heavy sheens [7 d] after the spill.” Two other pods did not experience unusual mortalities either in 1989 or later. Thus, this documented exposure does not appear to have been particularly substantial and appears to fall within the type of exposure that has been documented on several occasions, apparently without ill effects (Geraci 1990). The oil sheen that was present a week following the spill would have been highly weathered, and consequently, it is not clear how any whales could have experienced a significant and injurious exposure to VOCs at this time. Potentially killer whales might have consumed oil-contaminated food, salmon in the case of AB pod and oiled seals in the case of the AT1 group. Coho salmon, the main species eaten by PWS resident killer whales, did not have elevated hydrocarbon levels after the spill (Hom et al. 1996). These authors also noted that these migratory salmon attained most of their growth outside of Prince William Sound (and in 1989, before the spill). The AT1 transients might have consumed contaminated harbour seals. However, during the two to three months when harbour seals were most affected by the oil (Lowry et al. 1994) and clean-up was progressing, none of the AT1 transients was reported in the spill area. Although seals examined in the spill area in 1989 showed elevated hydrocarbon residues, and some seals were at least partially covered with oil, contaminant levels had returned to normal levels by 1990 (Frost et al. 1994). Thus, there is little evidence that killer whales were exposed substantial amounts of EVOS hydrocarbons through the food chain. Contact with oil is another effect pathway. Despite concerns that many whale biologists had before 1980 (Geraci and St. Aubin 1980), experiments and observations have filled in a number of blanks (Geraci 1990). The picture that emerges is that cetaceans appear to be well protected from potential injury from oil. The same mechanisms that serve to isolate and protect the internal environment of whales and dolphins from the sea water in which they swim also protect them from injury from contact with oil. It is important to appreciate that the heavy concentrations of thick, sticky material that characterize beach and nearshore conditions following a spill are absent from open water, once wind and waves act on the oil. Inhalation of high concentrations of

Page 5-46 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

VOCs is, perhaps, the most serious potential route of exposure. But high concentrations, which are present above the slick itself, are present for only a matter of hours after a spill (Fraker 2012; Hanna and Drivas 1993; Geraci 1990). Consumption of dangerous amounts of petroleum might occur under particular circumstances, but this has not been observed. The observation that killer whales were missing in the first two years following the EVOS, naturally raised suspicions that the spill was responsible. However, uncovering a plausible route of exposure has proven to be elusive; none of the suspected routes have proven credible. With the long periods during which the whales are unobserved, other factors might have been operating but were undetected. In the absence of more complete data on when, where, and how the killer whale deaths that were discovered in 1989 and 1990 were caused, conclusions must remain uncertain.

Discussion The discovery of a large number of deaths of killer whales following the EVOS naturally led to suspicions that the oil spill was involved. On the day of the spill, when the volatile organic compounds were at their highest concentrations in the air just above the surface, the whereabouts of the whales were unknown. However, even had they been within the comparatively small spill area that was affected in the first hours following the spill and would have potentially been exposed to the VOC maxima, the concentrations in the air would have been too low, by a full order of magnitude, to have caused serious injury. No killer whales were observed by surveyors during those first few days of the event, although a Los Angeles Times photographer did document some AT1 transients near the Exxon Valdez on day 2, about 36 h after the spill event began. The next sightings did not occur until one-week after the spill in lower Knight Island Passage, about 100 km from the spill site (Fraker 2012; Matkin et al. 1994, 2008). Despite a very active research effort during summer 1989 involving several vessels and aircraft, there was a long gap in killer whale sightings that extended from early April to late July (Matkin et al. 1994), when the killer whales were not detected in the spill area. The key question is how might the whales have been exposed to injurious amounts of VOCs, oil, or oil residues (e.g., in food or water). Fraker (2012) thoroughly reviewed the available information and concluded that the information does not point to a clear exposure pathway. VOCs were at their maxima (approximately 30 ppm) at some point in the first six hours of the spill (Fraker 2012; Hanna and Drivas 1993), but the threshold for causing neurologic effects would have been an order of magnitude greater (Geraci and St Aubin 1987). In addition, the whales would have had to have been in the small area of PWS that was oiled at that time. Cetaceans drink very little seawater, and thus, this would not offer a viable route of exposure. Oil residues in salmon, the primary food of the AB transients, were too small to have resulted in the ingestion of a harmful amount of oil. Although AB pod was seen in oiled waters in late March and again in early September, this was also true for four other pods, none of which suffered unusual mortalities.

July 2012 Page 5-47

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

It is possible that AT1 transients might have consumed harbor seals that used oiled haulouts during the first two to three months following the spill, when many seals were oiled. However, the killer whales were not seen in the spill area during this period. Except for the observation of AT1 killer whales on the day following the spill, there is no record of this group in oiled waters. As with resident killer whales, the AT1s would not have consumed enough water to have exposed themselves to toxic amounts of oil. Although the data set available for killer whales is remarkable, with each individual being accounted for in nearly every year, there are gaps of several months between sightings. Thus, the timing of a mortality cannot be known with an accuracy greater than the most recent observation. For example, the killer whales recorded missing seven days after the EVOS had not been observed since fall 1988, over six months earlier. Whether they died before or after the spill is unknown, as are the timing and causes of death. The six AB pod members that went missing over the winter of 1989-1990 clearly did die following the spill, but once again, the cause(s) of death is unknown. Complicating the picture further is the fact that at least four, and possibly more, of the whales that were discovered missing in 1989 had been shot during depredation incidents in 1985-1986 (Fraker 2012). There are a number of other confusing aspects to the situation presented by the killer whales: • If AB pod was affected by the spill, why was it the only resident pod so affected? What made this pod more vulnerable than the other? • If the AB mortalities detected in 1989 and 1990 were caused by the spill, did the same mechanisms that caused the early deaths (i.e., <1 week post-spill) also cause the later deaths (i.e., greater than 6 months post spill)? • If VOCs were involved in the deaths, why were the AT1s that were seen close to the Exxon Valdez on the day after the spill not affected immediately? • If ingestion of oil residues in food (i.e., coho salmon) was significant to the AB pod, why were other resident pods, which also rely on the same prey base, not affected (Matkin et al. 2008)? The situation with the AT1 transients differs fundamentally from that of the AB residents. The AT1s appear to be near the end of a long decline that began many years before the EVOS. In addition to the factors that limit the potential of any very small population from recovering, the AT1 transients, by virtue of their place high in the food chain, carry a very heavy burden of contaminants known to affect both reproduction and immune function. Whether the EVOS somehow contributed to the decline of ATI group cannot be known. What the data do tell us is that regardless of the cause(s) of the mortalities, AB pod is recovering as expected, based on its age and sex composition (Matkin et al. 2008).

Page 5-48 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Recovery of Killer Whales Regardless of the cause(s), killer whale mortalities did occur in PWS in the mid-late 1980s and 1990, and thus recovery from these losses can be considered. Because killer whales are long- lived, slow-reproducing animals, it is to be expected that recovery should be relatively slow. The EVOS Trustee Council (TC) has defined two recovery objectives, one for each killer whale form. The TC (2012a) will consider AB pod to be recovered when it returns from a post-spill low of 23 to the pre-spill abundance of 36. It numbered 28 in 2010, and thus is considered to be recovering, but not yet to have recovered (2012). Another sign of recovery is a normal reproduction rate, given the age and sex composition of the population (Matkin et al. 2008). The TC recovery objective for the AT1 group is a stable population trend (i.e., presumably an unchanging number). The AT1 group numbered just seven individuals in 2010 (TC 2012b), down from the 22 that were documented in 1984-1989. At the time of its discovery, the AT1 group comprised only 6 potentially mature females, and 10 mature males, which constituted the reproductively active part of the population. Because the AT1s are not known to interbreed with other killer whale groups (Barrett-Lennard 2000), inbreeding and other phenomena associated with very small population size (i.e., Allee Effects; Courchamp et al. 2008) appear to be operating, and consequently, a population as small as the AT1 group, had probably entered the “extinction vortex” (Gilpin and Soulé 1968) well before its discovery. The TC considers that the AT1 group is not recovering. As also noted by the TC (2012), the AT1s are carrying high concentrations of persistent organic pollutants that are in the range that has been linked with reproduction problems in other marine mammals (Ross and Birnbaum 2003, Ylitalo et al. 2001).

5.1.7.2 Pinnipeds

5.1.7.3 Harbour Seals

Exxon Valdez and Harbour Seals The population of Harbour Seals that inhabits PWS had been declining in size, since the early 1980s (Frost et al. 1994). Because of the impracticality of surveying all of the haulout sites in PWS to monitor the population status, the Alaska Department of Fish and Game established an aerial survey route to monitor the number of seals at a sample of the haulout sites. This monitoring program assumed that changes in the number of seals counted at these “index” sites reflected changes in the abundance of the population as a whole. Because some of the index sites were oiled during the EVOS and others were not, biologists believed that the effect of the EVOS could be estimated by comparing data from oiled and unoiled haulouts. From the survey data, Frost et al. (1994) estimated that at least 302 seals were missing from the predicted numbers and that it was likely that all of these “missing” seals had died owing to effects of the EVOS. To use data from the index sites, Frost et al. (1994) had to make several critical assumptions: 1) Seals at oiled and unoiled sites responded to changes in environmental conditions in similar ways over time, 2) seals have high fidelity to haulout sites, even under the adverse conditions resulting from the oil spill, (3) monitoring was adequate for detecting movements of seals to alternative

July 2012 Page 5-49

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments haulouts, and (4) sampling and analysis did not produce statistical biases. Hoover-Miller et al. (2001) found that every one of the above assumptions had been violated, which casts doubt on the conclusions about the effects of the EVOS. Of the 302 seals that Frost et al. (1994) believed had died as a consequence of the EVOS, only 14 were recovered, and 11 of these were pups (Spraker et al. 1994). Hoover-Miller et al. (2001) concluded that this number was too low to be consistent with the conclusion that 302 seals had died as a result of the EVOS. That is, the estimated 302 deaths reflected the violation of the assumptions on which Frost et al. (1994) had relied, and consequently, the actual number that died because of the EVOS was much lower. Of the dead seals, it is not clear how many were affected by oil exposure and how many died for other reasons – in most cases the cause of death could not be determined. Thus, in the case of harbour seals and the EVOS, there is disagreement about the nature of the effect. Frost et al. (1994) concluded that >300 seals died as a consequence, while Hoover-Miller et al. (2001) concluded that this number was an overestimate that resulted from flaws in the analysis conducted by the former group.

Exposure of Harbour Seals to Oil Unlike sea otters, which rely on fur for insulation and that, if fouled by oil, actively ingest oil during grooming, harbour seals rely on a thick layer of blubber and they do not groom themselves to remove oil that they have contacted. Thus, these two routes of exposure are not of great concern for hair seals. On the other hand, harbour seals do use shoreline habitats for hauling out and for pupping. And if these habitats are oiled, the mothers become contaminated each time they move across the oiled area. Consequently, pups are exposed to oil when they suck on their mothers for milk. These routes of exposure were demonstrated by Lowry et al. (1994). During the first few months of the EVOS event, harbour seals “… repeatedly swam through floating oil while feeding and moving to and from haulout sites” (Lowry et al. 1994). Although the spill occurred in late March 1989, about 6 weeks before the pupping season, and harbour seal rookeries were given priority for clean-up, not all of the rookeries were completely free from oil when pupping began. Thus, some pups and adults were exposed to oil for several weeks or more. Once harbour seal habitat was cleaned of spilled oil or the animals moved to unoiled areas, how much time might it take for them to become cleaned of residual oil? To answer this question, Lowry et al. (1994) soaked a heavily oiled piece of seal skin in sea water and found that in only seven days, it became clean enough that they doubted their ability to classify it as having been oiled. In confirmation of this, native subsistence hunters harvested two seals in an unoiled part of northwestern Prince William Sound, and only after retrieving them did the hunters determine that the seals had been oiled (Hoover-Miller et al. 2001). Furthermore, seals haulout to molt their fur in late summer, and at that time, in the absence of renewed exposure to oil, much of the remaining surface oil would have been removed from their bodies.

Page 5-50 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

The presence of hydrocarbon metabolites in the bile and of aromatic compounds in the blubber demonstrated that several of the seals examined after the EVOS had assimilated crude oil (Spraker et al. 1994). The lesions observed on the harbour seals were consistent with previous experimental and field observations of oiled seals. As with ringed seals that were experimentally exposed to crude oil (Geraci and Smith 1976), oiled harbour seals displayed conjunctivitis, which reversed when the seals were placed in clean water. There was also evidence of “minor and reversible” liver damage. Spraker et al. (1994) concluded that the most significant findings were lesions in the midbrain, which could account for the behavioural changes that were observed, particularly decreased flight distance, disorientation, and an increased tendency to haulout. As with the experimental observations of effects on ringed seals by Geraci and Smith (1976), the brain lesions were apparently reversible after the contamination source was removed. The compounds implicated by Spraker et al. (1994) were the highly volatile short-chain aromatic compounds, which rapidly dissipate with weathering. No neuronal lesions were discovered in seals collected in 1990, the year following the spill.

Discussion for Harbour Seals Harbour seals, which rely on blubber, rather than fur, for insulation, were affected by the EVOS to only a limited degree. The mortalities that were attributable to the spill were few and, as with the cetaceans, many of the mortalities probably owed to causes unrelated to the spill. The physical and neurological symptoms of exposure were in line with the experimental and observational results seen elsewhere in true (earless) seals and were reversible.

Recovery of Harbour Seals The Trustee Council (2012) set as a recovery objective for harbour seals that their numbers should be stable or increasing. Based on data collected from 1996-2005, this condition has been met. However, as discussed above, it is not clear that the harbour seal population in PWS actually suffered a substantial decline in size following the EVOS (Hoover-Miller et al. 2001). It is clear, however, that harbour seals in PWS were exposed to spilled EVOS oil in the shallow, nearshore environment and that the exposure resulted in certain associated, but reversible, effects, for example, conjunctivitis, liver, and brain damage. These effects were evident in the first summer following the spill, particularly soon after the event, but then the population recovered. By the summer following the spill, the symptoms had disappeared, as had the petroleum residues.

5.1.7.4 Steller Sea Lion Steller sea lions occur throughout much of the North Pacific Ocean, from southern California to (Loughlin et al. 1984, Schusterman 1981), and including Prince William Sound and adjacent waters, which were subject to the Exxon Valdez oil spill (Calkins et al. 1994). They are the largest of the eared seals (Otariidae). Two stocks are recognized: the Eastern Stock, which ranges from Cape Suckling, Alaska, south to southern California, and the Western Stock, which

July 2012 Page 5-51

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

ranges from Prince William Sound west to Japan. Steller sea lions are of conservation concern because of a broad-scale population decline and shift in population distribution, which has affected the Western Stock (Allen and Angliss 2011). The eastern stock population has been increasing since about the 1968, when fisheries related culls ended (DFO 2008). Steller sea lions spend a substantial amount of time hauled out on isolated beaches and rocky islets. Sea lions give birth at specific rookery locations from late May to July. Four rookeries (birthing sites) occur in the waters adjacent to PWS (none actually in PWS) and the Kodiak Archipelago and another 25 haulout sites are present in the same area. Although juveniles and adult males range widely among rookeries and haulouts, females show a high degree of site fidelity. Females usually produce a single pup, and about two weeks later they mate. After reaching the blastocyst stage, development of the embryo is suspended and implantation is delayed until about mid-October (Calkins et al. 1994, Schusterman 1981). Pups usually wean by their first year, although the dependency period can extend for three years.

Lack of Abundance and Mortalities of Steller Sea Lions in the EVOS The Western Stock of Steller sea lions was discovered to be in decline in the 1970s (Braham et al. 1980), and this decline continued at an average rate of 5.4 % into the early 2000s (Loughlin and York 2000). The count of adults and juveniles in 1989 was 18,135, down from 46,204 in 1976. Pup numbers were 4195 in 1989, down from 13,145 in 1979. However, from 2000 to 2004, the Western Stock increased by 5.5 % (Allen and Angliss 2011). The inflection point, from steep declines in the 1970s and 1980s to a moderate decline in the 1990s and eventual growth in the 2000s, appeared to coincide with a climatic regime shift in 1989 (Trites and Donnelly 2003). Regime shifts refer to changes in atmospheric and oceanic conditions (e.g., wind patterns, upwelling, sea surface temperature, etc.) that have cascading effects on ocean circulation, biological productivity, and energy transport through food webs (reviewed by Benson and Trites 2002). In effect, a regime shift could alter the carrying capacity of a regional ecosystem and ultimately the abundance of a particular species. The decline of Steller sea lions in Alaska, which was very much in evidence at the time of the EVOS (Calkins 1994), has been attributed to a dietary shift away from energy-rich, small schooling fish (e.g., herring, eulachon) and towards poorer quality gadids (e.g., walleye pollock) (Trites and Donnelly 2003). Chronic nutritional stress in the 1970s and 1980s was evidenced by relatively smaller and lighter bodies (up to 4.5% shorter and 22.5% lighter in 1980s than in 1970s) and relatively lower productivity (45% fewer pups in 1980s than 1970s) (Trites and Donnelly (2003). Over this same period, Steller seal lion scat analyses showed a reduction in small, energy rich schooling fish (from 61% to 20%) and increasing reliance on energy poor gadids (from 32% to 60%).

Page 5-52 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Exposure of Steller Sea Lions to Oil Sea lions were observed swimming in or near oil slicks, and oil was present near a number of haulout sites (Calkins et al. 1994). Moreover, tissue analyses indicated that the sea lions had been exposed to oil, although the values were highly variable (Calkins et al. 1994); some of the variability may have been the result of metabolism and elimination of the metabolites in the bile (Addison et al. 1986). Exposure may have come from absorption through the skin, inhalation, contact, or ingestion, either directly or by consuming contaminated prey (Engelhardt 1977). Numerous lesions were found in all of the major organ systems, except those of premature pups (Calkins et al. 1994). However, none of the lesions appeared to have been related to oil exposure. Although Steller sea lions did contact oil, it did not persist on them in the same way that it did on harbour seals (Calkins et al. 1994). The reason for this may have been the location and steep, high-energy slopes of the sea lion haulouts, which did not retain oil easily. Although some oil fouled Seal Rocks and Sugarloaf Island, only insignificant amounts of oil were seen at each rookery during late-June surveys in 1989; no oil was present in 1990. Also, oil from the EVOS that stranded on shorelines outside PWS was substantially weathered and patchy in its distribution (Gilfillan et al. 1995).

Discussion and Conclusions for Steller Sea Lions Calkins et al. (1994) concluded that none of the evidence provided conclusive evidence that the Exxon Valdez oil spill had an adverse effect on Steller sea lions, and consequently, there was no oil-spill effect from which to recover. The year of EVOS coincided with a reduction in the rate of Steller sea lion population decline, which had been occurring for at least 15 years prior to the spill and appeared to be linked to chronic malnutrition and a climatic regime shift (Trites and Donnelly 2003). As appears to be the case with the true seals, which includes harbour seals, and with cetaceans, it appears that sea lions appear to be relatively well protected from oil exposure. That is, sea lions do not rely on fur for insulation and their skin provides an effective barrier to oil absorption. Sea lion prey does not appear to consume or concentrate large amounts of petroleum from spills. Of course, they could be exposed to significant vapours from inhalation, but if that route of exposure occurs at all, it lasts only a matter of hours after a spill.

Summary for Steller Sea Lions Calkins et al. (1994) concluded that none of the evidence from the substantial study efforts provided conclusive evidence that the EVOS had an adverse effect on Steller sea lions, and consequently, there was no oil-spill effect from which to recover.

5.1.7.5 Sea Otters Unlike harbor seals, sea otters use fur for insulation and not blubber, which makes otters vulnerable hypothermia and possibly mortality if they fur becomes heavily oiled (Costa and Kooyman 1982; Riedman and Estes 1990). To maintain the air layer trapped in their fur, sea otters may spend up to 3 hours per day grooming (Costa and Kooyman 1982), which may also cause exposure to oil via ingestion. As a further consequence of the absence of a blubber layer,

July 2012 Page 5-53

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

sea otters must maintain their high metabolic rate by consuming approximately 25% of their body weight per day in invertebrate prey (Bodkin and Ballachey 1997); thus, ingestion of potentially contaminated food (i.e., benthic invertebrates) represents an important exposure pathway.

Acute Effects Sea otters were one of the most visibly affected marine mammals following the Exxon Valdez oil spill (EVOS). The estimated number of deaths from oil exposure was greater than 2,500 animals (CI 500 to 5,000; Garrott et al. 1993), although this estimate has been challenged (Garshelis and Estes 1997), and Garshelis and Johnson (2001) estimate 600 to 1,000 deaths. Following the EVOS, many oiled sea otters were transported to rehabilitation centres. The most common syndrome observed was shock, characterized by hypothermia, lethargy, and hemorrhagic diarrhea. In addition, many otters experience seizures at or near death (Lipscomb et al. 1994). Other common pathologies included anorexia, anemia, leukopenia, and azotemia. Histopathology revealed symptoms including interstitial pulmonary emphysema, gastric , hepatic lipidosis, and hepatic necrosis (Lipscomb et al. 1994).

Chronic Effects and Recovery A combination of evidence for continued elevated morality (Bodkin et al. 2002; Monson et al. 2000, 2011) and evidence for continuing exposure to oil (Bodkin et al. 2012) are required to attribute delayed recovery to an oil spill. The EVOSTC defines sea otter recovery as a return to pre-spill abundance and cessation of exposure to oil. To indicate that the population has failed to recover from the spill, evidence must be provided that sea otters continue to be exposed to oil (Bodkin et al. 2012). Between 1993 and 2000, the sea otter population in PWS had increased by approximately 600 animals to number 2 700 individuals (Bodkin et al. 2002). However, in heavily oiled areas of western PWS, sea otter population recovery rates were approximately half those expected following the spill, and in the most severely oiled areas, which were also those with the greatest sea otter mortality (e.g., northern Knight Island), no recovery was observed through 2000 (Bodkin et al. 2002). Several possible factors have been considered to explain this delay in recovery, including reduced survival rates (Monson et al. 2000, 2011; Ballachey et al. 2003), food limitation (Dean et al. 2002) and ongoing exposure to residual oil (Bodkin et al. 2012).

Food limitation Several indicators suggest that food limitation is not a factor in constraining the intrinsic recovery of the sea otter subpopulation at northern Knight Island. Between 1996 and 1998, the average prey energy available per unit mass of sea otter was four times higher at northern Knight Island than at an unoiled reference site, although this difference was not statistically significant (Dean et al. 2002). Despite this, the population at the unoiled site increased from 1996-1998, while the number of otters at northern Knight Island did not change (Dean et al. 2002). Further, while prey availability was variable and relatively low at both Knight Island and the unoiled

Page 5-54 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments reference site, foraging efficiency and the condition of young female otters was better at northern Knight Island (Dean et al. 2002). Finally, by 1998, birth rates did not differ between sites (Dean et al. 2002), and the survival of juvenile sea otters had returned to pre-spill levels, conditions which would be expected under non-food limited conditions (Monson et al. 2002). Thus, food availability did not appear to limit existing populations at either Knight Island or at the unoiled reference site. However, because this area of PWS has a long and continuous history of occupation by sea otters, food availability is substantially lower than would be expected at sites more recently occupied (Dean et al. 2002). Thus, while food availability is not likely to be the primary factor limiting the intrinsic recovery of the existing sea otter subpopulation at northern Knight Island, it is possible that relatively low food abundance in comparison to more recently occupied habitats limits growth by making this area less attractive to immigrants (Dean et al. 2002).

Exposure via foraging excavations In western PWS, clams comprise approximately 75% of sea otter diet (Dean et al. 2002), with foraging pit excavations ranging from >0.1 to 0.5 m deep. Bodkin et al. (2012) studied sea otter foraging activity and the possibility of encountering residual oil using observations of foraging pits, data from time-depth recorders to examine diving behaviour (e.g., depth), and published estimates of remaining oil (Short et al. 2006; Boehm et al. 2008). They concluded that sea otters foraging at northern Knight Island would encounter remaining Exxon Valdez oil an average of 10 times a year, with females 2.5 times more likely to encounter remaining oil than males. Further, instances of intertidal dives increased in the late spring and early summer, when most adult females were giving birth, indicating a pathway of exposure not only for adult otters, but potentially for dependent pups as well (Bodkin et al. 2012). Conversely, Boehm et al. (2011) suggested that there was clear spatial separation between areas where subsurface oil residue was present and areas where sea otters would be expected to forage. They contend that sea otter foraging pits are rarely more than 15 cm deep (shallower than buried oil residue), and that sea otters would not dig foraging pits in the middle and upper intertidal sediments (in 2005 the highest foraging pit on shore was at +0.4 m, and in 2006 the highest pit was at +0.85 m) (Boehm et al. 2011). Therefore, they report that sea otters would not be expected to encounter or be exposed to oil residue. However, these contentions were refuted by Bodkin et al. (2012), using data from the time-depth recorders implanted on 19 sea otters. They demonstrated that foraging dives ranged from +2.7 to -92 m below sea level (MLLW), with dives in the intertidal zone accounting for 5 to 38% of all foraging dives. Further, of the 16 050 intertidal dives per year made by female otters, 18% occurred above the +0.80 m tidal elevation. Males made 4 100 intertidal dives per year, with 26% of these above the +0.80 m tidal elevation (Bodkin et al. 2012). Sediment samples were also collected directly from 41 sea otter foraging pits on four beach segments (Bodkin et al. 2012). Analysis indicated that eighteen of these contained oil residue, with total PAH concentrations at oiled Disk Island exceeding 56 000 ppb (concentrations from this area reported by Boehm et al. (2011) ranged from 3-325 ppb). As Boehm et al. (2011) point out, a combination of biological and physical requirements must be met to create optimal

July 2012 Page 5-55

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

foraging sites. However, disagreement exists as to whether these requirements overlap with the occurrence of residual oil on the shores of PWS.

Reduced survival/Population sink From 1989 to 1998, survival rates in the western PWS sea otter population were reduced relative to pre-spill survival rates (Monson et al. 2000). The magnitude of the decrease was dependent on age class and time since the spill, with greatest reductions seen in the survival of younger age classes (<9 years) the year after the spill, with the survival of older adults apparently less affected. However, the survival of younger age classes increased with time, while the survival of older animals decreased (Monson et al. 2000). To explain this observation, it has been suggested that in the first few years after the spill, exposure levels in the sink population were relatively high, and that pups and juveniles alive during or born right after spill were unable to cope with the exposure (Monson et al. 2011). In contrast, decreasing exposure levels led to improved juvenile survival, while otters born into the sink population may have experienced chronic, low- level oil exposure during development, leading to decreased survival in adulthood (Monson et al. 2011). In contrast, Garshelis and Johnson (2001) found that sea otter populations in western PWS were as high as or higher during the 7-year period after the spill as they were during the mid-1980s. Their results indicated that pup production was abnormally high in both 1993 and 1994, returning to pre-spill levels in 1996 (Garshelis and Johnson 2001). The authors attribute this increase to an enhanced food supply in the early 1990s, possibly as a result of ongoing recovery from the uplift caused by the 1964 (magnitude 9.2; Garshelis and Johnson 2001). It is difficult to use this data to examine population trends on an individual site basis, as complete surveys were carried out only in 1991 and 1996. However, for the Sound as a whole, the authors found that between 1991 and 1996, the overall PWS sea otter population increased by 13%, an annual increase of 2.5% (Garshelis and Johnson 2001). This finding is in in good agreement with more recent work by Bodkin et al. (2011), who found that the overall population trend in western Prince William Sound as a whole is one of slow, steady growth, with increases from 2 150 animals in 1990 to approximately 3 000 animals in 2009, and annual rate of increase of approximately 2.6% (Bodkin et al. 2011). Conversely, the subpopulation at northern Knight Island appears to have remained depressed (Monson et al. 2011). The authors hypothesize that the northern Knight Island subpopulation is acting as a population ‘sink’, with the cumulative loss of approximately 900 animals since 1990 (Monson et al. 2011). The authors used the age distributions of living and dead animals, as well as estimates of sea otter population size, to predict the number of otters in the sink population, as well as the number lost to this sink. Model results indicate that the sink population remained steady at just over 900 animals between 1990 and 2009, with prime-age survival remaining at 2- 6% below pre-spill levels. Further, this effect did not appear to be dissipating, with modeled estimates of animals lost annually remaining nearly constant from 1994 to 2009 (Monson et al. 2011). However, results of comparisons between predicted and observed sea otter populations suggest some reason for optimism: the most recent population surveys indicate that sometime between 2007 and 2009, the northern Knight Island subpopulation experienced positive

Page 5-56 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

population growth (while the model indicated that it should still be declining) (Monson et al. 2011).

Summary for Sea Otters There remain disagreements in the peer-reviewed literature with respect to the recovery of sea otters from the EVOS. However, researchers appear to agree on some key points: • The spill caused substantial acute mortality in the PWS and wider Gulf of Alaska sea otter populations; • Residual oil remains in PWS (although debate continues as to the likelihood of exposure to this oil); and • The overall sea otter population in PWS has been slowly and steadily increasing since the early 1990s, with birth rate and juvenile otter survival returning to pre-spill levels by 1998. While populations in the rest of the Sound recovered, the northern Knight Island sub-population appeared to experience prolonged effects of the spill. However, recent work has indicated that this population may also be beginning to recover. Thus, while recovery for PWS as a whole appeared to occur within 10 years, populations inhabiting heavily oiled areas may require closer to 20 years before positive population growth occurs.

5.2 Human Environments Associated with the Marine Environment The potential effects of marine oil spills on the human environment and the time required for recovery are dependent on the volume, location, the nature of the resources affected, the timing and intensity of traditional and non-traditional activities in the area, and the duration of clean-up and recovery. A review of the literature related to the socio-economic effects of marine spills indicates that information is available for only a few of the major marine spills that have occurred, and much of the published information focuses on a detailed examination of EVOS, although this event occurred more than 20 years ago. There are a number of reasons that explain the limited amount of recent information on the socio-economic effects of marine spills. First, the number of such spill events and the amount of oil being spilled into the marine environment has been decreasing over time. Statistics from the International Tanker Owners Pollution Federation Limited (ITOPF, 2011) show that both the number of large spills (those exceeding 7,000 tonnes of oil) and the total amounts of oil spilled have been steadily declining since the 1970s (Figure 5.8). The number of large spills has dropped from an average of 24.5 per year in the 1970s, to 9.3 per year in the 1980s, to 7.8 per year in the 1990s, and to 3.3 per year in the 2000s. Thus, over time there have been less frequent large spills for which to study the effects on the human environment and this can be attributed to improvements in vessel design and navigational technology and practices.

July 2012 Page 5-57

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Source: ITOPF 2011 Figure 5.8 Marine Oil Spill Statistics 1970 to 2011 Furthermore, Figure 5.9 shows that, in those years when large volumes of oil were spilled into the marine environment, large quantities of oil came from one or two large spills. There has not been a large marine spill event (greater than 200,000 tonnes) since 1991 and, with the exception of the Prestige spill, there has not been a large marine spill since 1996.

Source: ITOPF 2011 Figure 5.9 Major Marine Oil Spill Events 1970 to 2011

Page 5-58 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Second, very few large marine oil spills have occurred in coldwater environments like the north coast of British Columbia. Figure 5.10 shows the location of larger marine oil spills and only two, EVOS (37,000 tonnes) and Braer (85,000 tonnes), occurred in coldwater marine environments. Other spills in coldwater environments, which are described earlier in this review, are smaller spills.

Source: ITOPF 2011 Figure 5.10 Location of Major Marine Oil Spill Events Thus, there are few examples of marine oil spills that represent a situation analogous to the British Columbia coast. For this reason, much of the assessment of the effects of a marine oil spill on the human environment is based on published reports related to the EVOS spill. In reviewing this information, it must be remembered that EVOS occurred in 1989. Since then, much has been learned about how to avoid such incidents in the first place, as evidenced by the declining number of major spills, and how to contain, clean-up and rehabilitate affected environments. Were a similar spill to occur today, it is likely that there would be similar types of effects on the human environment but the magnitude and duration of the effects would likely be reduced. These reduced effects have come about due to the increased collective learning and enhanced spill response training strategies that have occurred in North America since 1989. This conclusion is based on the experience from the Selendang Ayu spill which occurred off the coast of Alaska in December 2004. This involved a spill of about 7,640 barrels (1,215 cubic metres) of intermediate fuel oil and diesel oil from a bulk carrier (soy-beans) bound for . A comparison of the socio-economic effects from the EVOS and the Selendang Ayu spill demonstrates the extent to which response capacity has improved since 1989, with effects of the Selendang Ayu spill providing a more accurate description of potential effects on the human environment were a similar spill to occur now. An in-depth review of the effects of the EVOS and Selendang Ayu spill on the human environment, as well as measures taken to reduce effects, is provided in Appendix C.

July 2012 Page 5-59

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

5.2.1 Commercial Fishing Following EVOS, fisheries for salmon, herring, crab, shrimp, rockfish and sablefish were closed in 1989 throughout Prince William Sound, Cook Inlet, the outer Kenai coast, Kodiak and the Alaska Peninsula. The commercial fisheries for shrimp and salmon remained closed in some parts of Prince William Sound through 1990. As a result, commercial fishing activities in Prince William Sound and the Gulf of Alaska were disrupted and the incomes of commercial fishers declined. According to the Exxon Valdez Oil Spill Trustee Council (EVOSTC, 1994), the recovery objective for the commercial fishery was originally based on the recovery of “population levels and distribution of injured or replacement fish used by the commercial fishing industry”. This test proved problematic, however, because it essentially meant that commercial fish population numbers and distributions would have to return to levels that would have existed in the absence of EVOS, or, as a substitute measure, to reach pre-spill levels (Integral Consulting, 2006). The issue was that harvesting activities and incomes from commercial fishing normally fluctuate as a result of changing fish populations and a variety of other factors, so it may never be possible to achieve “recovery” if fish harvests and incomes in the base year before the spill event were unusually high. For example, it was noted that 1987 and 1988 were years of relative prosperity for Alaskan fishermen because of high salmon prices so the revenues lost as a result of EVOS were exacerbated by expectations for continued high revenues. As a result, the recovery objective for commercial fishing was redefined in 2001 to be “when the commercially important fish species have recovered and opportunities to catch these species are not lost or reduced because of the effects of the oil spill” (Integral Consulting, 2006). The largest effects on commercial fisheries were related to declines in pink salmon and Pacific herring, so restoration efforts focused on these species (See Sections 5.1.6 and 5.1.5). By 2001, both pink and sockeye salmon had been removed from the list of injured species. However, herring populations have still not returned to pre-spill conditions due to a number of factors not related to the EVOS (Section 5.1.5.1); as a result, the herring fishery in Prince William Sound was closed for 15 of the 21 years after the spill. These factors include disease aggravated by high population density, food scarcity, and poor ocean conditions (Integral Consulting, 2006). More recent studies conclude that the 1993 herring collapse was due to poor nutrition and disease (Rice and Carls 2007; Deriso et al. 2008; Hulson et al. 2008; Pearson et al. 2011). As a result of diminished herring populations, the EVOSTC (2010) still considers commercial fishing to be “recovering” from the effects of the spill. It should be noted though, that there have been no non- herring, spill-related, district-wide fishery closures related to oil contamination since 1989. Although the entire commercial fishery is still considered to be “recovering”, many of the individual components of the fishery are considered to have recovered. For example, the review by Integral Consulting (2006) observed that salmon harvests in Prince William Sound in 2005 were the largest on record and nearly twice the 1987 harvest (see Figure 5.11). However, the revenues associated with salmon harvests did not increase proportionately due to various reasons, including the increasing availability of farmed fish. Thus, the commercial aspects of the salmon fishery was considered to have fully recovered within 12 years for all salmon species,

Page 5-60 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

although the total salmon harvest had reached pre-spill (1988) levels in 1994 (five years after the spill), following poor years in 1992 and 1993.

Source: Ashe et al. (2006) as reported in Integral Consulting (2006) Figure 5.11 Annual Numbers of Salmon Harvested by Commercial Fishermen in Prince William Sound from 1976 to 2005 While disruptions to commercial fishing still exist today, many of the observed changes in income are not directly attributable to EVOS (EVOSTC, 2010). Factors affecting commercial fishing in Prince William Sound and other parts of Alaska include: • changes in the commercial fishing industry, including the world supply of salmon (due primarily to farmed salmonids) and corresponding reduced prices • entry restrictions in certain fisheries (such as Individual Fishing Quotas, for halibut and sablefish) • allocation changes (e.g., a reduction in the allocation of Cook Inlet sockeye salmon to commercial fishermen) • reduction in processing capacity • spatial limitations of groundfish fisheries in the spill areas because of ongoing sea lion management protocols These factors make it difficult to identify spill-related effects on the commercial fishery. Recovery from the Selendang Ayu spill followed a similar pattern. The community most affected by the spill (Dutch Harbor/Unalaska) is heavily reliant on commercial fisheries, processing more than $1 billion per year of fishery resources, with the majority of this coming from groundfish and crab (Kelty 2005). Following the spill, the Alaska Department of Fish and Game closed fishing in the area because of the amount of oil spilled in the area and the contamination found in various samples of fish and crab species (Kelty 2005). Areas were reopened to fishing in October, 2005, about 10 months after the spill (Alaska Department of Fish and Game 2005). The

July 2012 Page 5-61

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments closure resulted in a lost opportunity to harvest 175,000 pounds of crab, Pacific cod, and halibut. The resulting damages for the local small boat fleet were expected to be partially recovered through insurance payouts (Kelty 2005). A working group was established to address the risk from the spill for commercial fisheries. It implemented a fisheries water-quality sampling program to monitor water and seafood in support of the State of Alaska’s “zero tolerance” policy, which aims to prevent contaminated finfish and shellfish species from reaching the consumer when an oil spill occurs (Pearson, 2005). Water quality sampling and seafood safety inspections were conducted for two crab fisheries and important pollock, Pacific cod, and halibut fisheries. There is some information on the effects of other marine oil spills on commercial fisheries. Knapp (2005) observed that the 2003 Tasman Spirit spill (28,000 tons) off the coast of led to a sharp decline in the sale of seafood in nearby city markets, but did not affect export markets because buyers were aware that the fishing grounds were not directly affected by the spill. The 2002 spill of 30,000 tons of oil from the Prestige off the Atlantic coast of had a negative effect on French oyster sales in 2003 but the market recovered in 2004, although much of the decline was subsequently tied to the depressed European economic climate and increasing imports from Chile. Thus, marine spill effects on commercial fishing appear to be in the range of one to several seasons and the extent of losses as a result of the spill are often difficult to isolate from other trends and events that are affecting commercial fishing in the surrounding region.

5.2.2 Traditional Use

From our literature review of the socio-economic effects of marine spills, the available data are clearly skewed in favor of a few large spill events, particularly the EVOS. In part because of the magnitude of the spill, the high profile and often sensational news coverage and extended litigation, and the cultural, social and regulatory context for oil spill mitigation (a context that is quite different from what exists in Alaska today), the EVOS event is now the lens through which many social scientists view marine spill effects on aboriginal traditional use of marine areas and resources. In an attempt to clarify some points that are contentious, unsubstantiated by fact or debatable from a scientific point of view, this section focuses on the effects on and recovery from EVOS on traditional use of subsistence resources and country foods in the Alutiiq region (Fall et al. 2001, 2006). To further illustrate potential effects and recovery, this review also examines the effects of the Selendang Ayu spill on traditional use of country food resources on Unalaska Island in the Bering Sea area. This includes a discussion of a much changed regulatory framework that makes oil spill mitigation more efficient, more effective, and timelier through modern response capabilities than was ever conceived of at the time of the EVOS. The importance of country food/subsistence/traditional resource use and sharing to Alaska Natives is well documented. Therefore, marine oil spills have the potential to affect both subsistence and country food resources and the human environment in which they are harvested, processed, shared and consumed. The phrase “country food,” refers to wild fish, game, and botanical resources, whether or not harvested within the legal contexts of subsistence, personal

Page 5-62 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

use, or procured via informal or formal markets. It remains difficult to distinguish among and between specific effects of marine oil spills on traditional use, especially through systematic analysis of subsistence harvest data, although these data provide a good perspective on trends through time and in spatially explicit ways that show, how, when and where recovery has happened and is happening. One of the frequently ignored problems in this type of work is that perceived correlation often does not clarify or establish in a scientific way that proposed causes act through a plausible mechanism to the perceived effects. Some of the key food system concerns related to marine oil spill effects on the human environment are: (1) the direct effects of oil spills and contaminants pollution on the quantity and availability of country food resources, (2) habitat disruption associated with marine spills, (3) access limitations for country food harvesting in or around spill areas, (4) competition for country food resources caused by reductions in availability or other spill-related factors, (5) disruption/displacement of migration routes and feeding areas for marine mammals as a result of spills and clean-up activity, (6) fear of contaminated resources. The definition of “recovery” from EVOS concerning subsistence, as defined and used by the Trustee Council is: “Subsistence will have recovered when injured resources used for subsistence are healthy and productive and exist at pre-spill levels. In addition, there is recognition that people must be confident that the resources are safe to eat and that the cultural values provided by gathering, preparing, and sharing food need to be reintegrated into community life.”. http://www.evostc.state.ak.us/recovery/status_human_subsistence.cfm In addition to the direct, even immediate and perhaps short-term effects of the loss of natural resources available for harvest because of the initial spill effects, other human environmental factors can affect spill recovery. These include food safety concerns and demographic, social, cultural, and socioeconomic change. These latter changes need to be considered as well as long- term effects associated with climate change and ocean warming, including coastal weather events such as storm surges and changes in wind direction, extreme weather that affects nutrient upwelling, and changes in the distribution of marine species, interannual variability and seasonal change. In addition, traditional use is also affected by increased pressures on alternative wild resources; increased competition for subsistence resources by sport fishing charters, deer hunters, and natural predators (such as sea otters and orcas, to mention two examples); and increased awareness of the prevalence and dangers of PSP and other contaminants from industrial and military resources, not related to oil spills. Other issues affect the socio-economic aspects of recovery and traditional use activities. These can include opportunities for wage employment; spill restoration fund availability/land sales; prolonged litigation as a long-term spill effect; the revitalization of subsistence uses; and support for communities to rebuild, restore, and develop self-reliance within the context of traditional/historic lifeways. While simultaneously coping with a spill, aboriginal communities must also deal with potentially positive benefits of a large cash infusion that supports traditional harvesting and sharing activities, especially in many rural communities where employment is limited and the harvest costs are extremely high. High intergenerational change is also affecting

July 2012 Page 5-63

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

who harvests and where they harvest, resulting in situations where the Elders are most active on the land.

5.2.2.1 Recovery of Subsistence Harvesting after the EVOS According to the Trustee Council’s summary (much of it relying on ADF&G Division of Subsistence study data), in the first year after the spill, subsistence harvest declined substantially in 10 villages within Prince William Sound, Cook Inlet and Kodiak. The villages of Tatitlek and Chenega (the Alutiiq villages in PWS closest to the spill) reduced their harvest by 56 and 57 percent, respectively. Smaller reductions occurred in villages outside of the Sound. The primary reason for these reductions provided by the harvesters was the fear that oil had contaminated the resources and made them unfit to eat. Other factors including work on clean-up jobs and food provided by outside sources were also contributing factors in a decline in subsistence harvest activities. Harvest levels have generally increased in many communities since the spill, but results of harvest surveys are as always variable. By 2003, harvest levels were generally higher than pre- spill levels in the communities in Cook Inlet, but lower in Kodiak and Prince William Sound, with the exception of Cordova. As a point of context, the situation in Cordova, a mostly non- Native commercial fishing town, is not portable to other rural communities, especially those comprised primarily of Alaska Native residents and where subsistence is critical to the food security component of the system. Even though the harvest levels in the PWS communities were not as high as pre-spill estimates, they were within the range of other Alaska rural communities. Harvest composition was also reportedly altered by the spill. In the first few years following the spill, people harvested more fish and shellfish than marine mammals, in part because of the reduced number of marine mammals and the perception that these resources were contaminated and unsafe to eat (EVOS Trustee Council). EVOS Trustee Council lists subsistence use as “recovering” and harvest levels from villages in the spill area are comparable to other Alaskan communities. Fall et al (2006:393) provided a summary in their comprehensive 2003 survey that addresses recovery on the basis of three questions. 1) Are the resources healthy and are populations at pre-spill levels? Data supporting a recovery of subsistence use include a rebound in subsistence harvests to pre- spill levels, the diversity of harvests matches or exceeds pre-spill harvest levels, and most people in the communities use wild resources and participate in harvesting them. In contrast, evidence that subsistence uses have not fully recovered include the reduced availability of herring, harbor seals and clams/intertidal resources; lower harvests of clams and seals in several communities than before the spill, the need for some people to go farther to harvest resources, and other reductions in harvests compared to pre-spill levels at the household level.

Page 5-64 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

2) Are people confident that the resources are safe to eat? For this question, most respondents who offered an opinion said that chitons, herring and harbor seals were safe to eat, which would be evidence that recovery has occurred. Whether such a recovery claim is an artifact of cultural perception or an ecological fact is still a research question in progress. The evidence Fall cites that there are continuing food safety concerns includes low and declining confidence in the safety of eating clams in some communities; some people blame EVOS for lower harvest rates of shellfish by claiming EVOS conditions gave rise to the growing threat of paralytic shellfish poisoning in the region (other communities in PWS blame climate change and changes in weather and seasonality on paralytic shellfish poisoning, without reference to the EVOS.). 3) Have the cultural values connected to subsistence uses been reintegrated into community life? The evidence that subsistence values are recovering based on cultural values is that resource sharing, a cultural hallmark by most standards, is frequent, widespread and involves most households in the region. In seven of the communities, most respondents reported that young people were learning subsistence skills, although there are still intergenerational issues with respect to motivation, desire to participate and ability to participate. According to Fall, the evidence that values are not recovering included the fact that many survey respondents claimed youth were not learning enough about subsistence skills, many said that the elders’ influence is declining, and most respondents reported that their cultural way of life had not recovered from the effects of the spill. However, the challenge of discerning cause and effect is relevant here because of the direct, indirect and cumulative effects of cultural, social and economic change. Some of these may or may not be related to the oil spill, and while others are general societal drivers and are clearly not directly related. As noted, it is difficult to distinguish the specific effects of marine oil spills by analyzing subsistence harvest data in isolation, with shellfish an excellent example. The perception that there is a link between increased PSP risk and EVOS is a concept that may or may not be captured in the ADFG harvest data, but it is still a substantial local concern. Subsistence harvest data provides a broad look at broad trends that, in the EVOS case, indicate recovery has been and is occurring in the human environment. Fall (2009) also conducted a comprehensive assessment of the effect of the EVOS on (primarily Alaska Native) subsistence users in “Long Term Consequences of the Exxon Valdez spill for Subsistence Users of Fish and Wildlife.” In it he describes a number of sociocultural “core studies” that were conducted in communities in the EVOS area, specifically Cordova/Eyak and Seward, in the 1970s long before the Exxon Valdez spill. The State of Alaska ADF&G Division of Subsistence had conducted baseline studies in the 15 predominantly Alaska Native communities in the EVOS area, plus the four larger mixed communities, although baseline here can mean one observation or set of observations at one point in time, within one season, or across a very limited number of seasons.

July 2012 Page 5-65

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

These studies look at demography, employment patterns and quantified data regarding the dimensions of the subsistence harvests. These data were collected by State and Federal government agencies with responsibility to assess the potential impact of oil development on its citizens. After the Exxon Valdez spill, in 1990, the Division of Subsistence conducted systematic household surveys in the 15 spill area communities pertaining to the first spill year (1989). In 1991 they conducted interviews in seven of the 15 communities for the second post-spill year (1990). Subsistence harvests declined substantially after EVOS, at least by comparison to the pre-spill years. People were suspicious of the quality of the wild resources and feared the harvests had effectively been poisoned. By the second year, some rebound had occurred in the Kodiak and Lower Cook Inlet areas, but not in Chenega and Tatitlek, the Alutiiq villages closest to the spill. In addition to taking advantage of cash wages available for working on spill clean-up, other reasons, primarily contamination concerns, were cited for the decline in subsistence harvests. An “Oil Spill Health Task Force” consisting of state, federal, Alaska Native organizations and Exxon was established to address the possibility of contaminant transfer through the food chain to humans. The Oil Spill Heath Task Force was formed in 1989 to assess the health of the foods by analyzing fish and shellfish from the region. Finfish were found to be safe to eat (with some caveats), but shellfish and crabs from highly contaminated beaches were not. People continued to express fear over the health of wild foods and it took another three years of testing and agency dissemination of results to allow most people to feel comfortable enough to conclude that wild foods were safe to eat. Immediately after the spill, emergency food relief programs in which villagers from non-spill areas shared portions of their subsistence harvests (with logistics assistance by the State of Alaska and Exxon) provided most households with subsistence food. Other fish and grocery distribution programs also provided short-term compensation for the loss of subsistence foods. These programs were not a perfect solution, but enabled households to continue to function. Meanwhile, incomes increased during the oil spill year (1989) for all household types due to increased wage incomes. Clean-up jobs provided income above normal levels, and the active elder households (normally low income producers) actually earned the most income during the oil clean-up phase. The spill presented conditions unfamiliar to the subsistence users, resulting in households acting with great caution because doubts persisted about the effects of the spill on food quality, safety and the health of marine, country foods. There were negative effects related to the lack of resources available to share within communities and concerns over harvest quality, but the basic fabric of society remained unchanged (Fall et al. 2001:2). In addition to increased employment opportunities immediately after the spill, the EVOS Trustee Council provided compensation to Alutiiq villages for community-based projects during the years following the Exxon Valdez oil spill. This helped to counter the injuries to the natural

Page 5-66 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

environment that these communities relied upon (Fall et al. 2001). Projects included fish stock enhancement, subsistence and educational facilities, cultural education projects, wild foods safety projects, mariculture development projects, wild resource assessments and local participation in restoration projects. Such community-focused projects reduced the long term effects of the spill, and enhanced the subsistence economy and cultural values shared by the communities. Economic and cultural revitalization within the Alutiiq villages was a positive development associated with EVOS. Tribes benefitted from projects designed to restore the natural environment. After the $900 million settlement of natural resource damage claims against Exxon by the state and federal governments, the Trustee Council administered a restoration fund. The Trustees considered subsistence to be an injured resource service, and the Trustee Council directed funds toward subsistence restoration projects, resource enhancement and cultural revitalization. Over 450,000 acres of mostly Alaska Native Corporation lands were voluntarily sold into the Council’s habitat protection program. However, new sources of cash also provided the means for people to leave the villages, raising uncertainties about the future of some communities. In addressing the issue of recovery in the human environment, Fall et al. (2006) note that no complete recovery of the human environment can occur because a return to 1989 conditions is impossible. Additionally, the social landscape is complex and, as noted earlier, a broad suite of factors affecting the human environment continues to change. Some of the changes in the natural and social environments are seen by local residents as being caused by EVOS, but the link for other changes is not so clear. Fall et al. (2006) state that “recovery will have occurred when the people of these communities believe that their communities will have a strong and viable future that builds upon their past, a future that they themselves must help to shape.” Given the terms of this definition, recovery in the human environment of the Alutiiq region is on-going, and probably not yet fully recovered relative at least to pre-spill conditions. However as Fall et al. (2001:287) concludes elsewhere in a discussion of EVOS and Alaskan oil development, the basic organization of the factors of production and distribution remained stable in the villages. The spill did not trigger a collapse at the basic local level of the extended household networks typical of the Alutiiq villages. “While the spill created major local disruptions of food procurement and employment patterns, the spill did not transform the pattern of relationships in the subsistence sector. The traditional extended kinship networks adapted to the short-term crisis of food production and distribution at the local level without major dislocations in the underlying structure of production and distribution.” Fall et al. (2001:292) also concluded “While the injuries have been great, and tragic to some users (such as the Alutiiq), they have not been considered irreparable . . .” The pattern of relationships within the villages remained intact partly because resources were available to compensate for lost subsistence harvests and households shared them as they typically shared harvested natural resources. In this broad sense, recovery is occurring in the human environment.

July 2012 Page 5-67

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

5.2.2.2 Selendang Ayu Harvest Effects In this case Alaska Native communities addressed concerns related to oil in the environment (spilled Bunker C oil) with government scientists and industry experts. During the Selendang Ayu response, Qawalangan tribal members and government agency personnel cooperatively addressed aboriginal concerns with representatives of the responsible party. Those collaborative studies (summarized in Mauseth et al. 2008) describe risk-based public health evaluations of actual spill effects. They objectively evaluated the public health risk of subsistence food consumption including addressing the various perceived and real risks posed by the spill. ADF&G Division of Subsistence did not conduct household-level data documenting subsistence harvest and use after the Selendang Ayu spill as they did for the EVOS response; therefore there are no detailed data on the extent to which the spill resulted in changes in subsistence harvest and use. The Selendang Ayu spill was orders of magnitude smaller than EVOS and food safety concerns were quickly and cooperatively assessed, resulting in a very different subsistence effect scenario than EVOS. Despite the assurances the foods were safe to eat, some people still remained uncertain about the food quality and retained a more skeptical outlook.

Table 5.10 Summary Table for the Selendang Ayu Oil Spill Oil Spill Name Selendang Ayu Location Alaska, USA Year 2004 Oil Type No. 6 Fuel Oil Specific Name Bunker C Volume (metric tonnes) 1,072 Platform Tanker Environments Affected Marine Valued Ecological Component (VEC) Studied Birds Summary: On December 8, 2004 the tanker vessel Selendang Ayu ran aground off the coast of Unalaska in the Aleutian Islands. Approximately 112 km of shoreline were oiled with 30 km requiring cleanup. The estimated volume of recovery from the initial cleanup efforts could not be found in the literature. The initial cleanup response went from December to February 2005. Most of the cleanup was complete in the spring with some isolated areas still being cleaned in the summer. Studies of birds were followed sufficiently for a year and half after the spill to report on the status of recovery. After the food safety project was complete, the Qawalangin Tribe of Unalaska submitted a claim for damages associated with loss of subsistence use of natural resources as a result of the Selendang Ayu spill. The claim was denied by the U.S. National Pollutions Funds Centre for various reasons, notably that the Tribe did not clearly identify the natural resources used for subsistence purposes that were lost as a result of the spill (NPFC, 2009).

Page 5-68 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

5.2.2.3 Selendang Ayu Summary of Effects on Traditional Use In reviewing the literature, one of the challenges in assessing the potential effects of a marine oil spill on subsistence harvesting is the lack of comparable community-by-community pre-spill harvest data. While governments typically report commercial and recreational fishing information at a regional or sub-regional basis, there is rarely detailed harvest data available, with the exception of the ADF&G Division of Subsistence studies. In the absence of detailed harvest data, communities and researchers are forced to describe pre-spill harvesting activities in qualitative terms after a spill has actually occurred. This can lead to difficulties when using qualitative data as the basis for compensation settlements. In the Selendang Ayu spill, there was a minimal spill effect that was assessed in relation to the subsistence resources used traditionally in the area, and recovery of those resources occurred apparently with no long-lasting effects on the human environment.

5.2.2.4 Selendang Ayu Subsistence Food Safety and Recovery Similar concerns about food safety and subsistence use arose following the Selendang Ayu spill. Although a study on subsistence food consumption safety was issued in 2006, 16 months after the spill, stated that levels of PAHs were not present at levels of concern for human health (Department of Health and Social Services and U.S. Fish and Wildlife Services, 2006), subsistence harvesters were cautioned to avoid harvesting or eating animals that were evidently oiled or were found on oiled beaches, or that were behaving oddly or have unusual lesions. However, residents of Unalaska were already concerned about other types of contamination (PCBs, oil from previous spills and ship disposal, and discharges of sewage and effluent from seafood processing) and were choosing to practice subsistence harvesting at other locations (Kohout and Meade, 2008). Shortly after the spill, the Selendang Ayu Subsistence Fishery Advisory Group developed a subsistence food sampling plan. Fifteen different areas were sampled for a variety of subsistence foods and analyzed for PAHs (Arnold 2006, Appendix A). During June and July of 2005, composite samples (10 to 20 individual organisms) of blue mussels, black chitons, and green sea urchins were collected from Unalaska Bay in areas frequented by residents of Unalaska and Dutch Harbor. During August and September of 2005, after oil spill clean-up activities were essentially completed for the summer, composite samples of blue mussels, black chitons, and green sea urchins were collected near the Selendang Ayu grounding and spill area (i.e., Anderson, Cannery, Kashega, Kismaliuk, Makushin, and Skan Bays). Also, twenty-three known frequent subsistence food consumers were interviewed about their seafood consumption in late September and early October of 2005. Households were asked how much of a resource they harvested per year. The harvest data was analyzed along with other ADF&G Division of Subsistence data to estimate the total annual harvest for the Unalaska/Dutch Harbor community (Mauseth et al 2008).

July 2012 Page 5-69

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Arnold concluded that PAHs in subsistence resources from Unalaska Bay are not present at levels of health concern. One chiton sample collected from Summer Bay contained PAHs above risk-based screening criteria; however, PAHs were below risk-based screening criteria in three other composite samples collected from the same area. The source of PAHs was from another source, but not Selendang Ayu oil. As expected, PAHs were highest near the spill site, although, for the samples evaluated, the concentrations were not at a level to present a public health concern (Arnold 2006). The study concluded that the levels of PAHs in subsistence food resources were expected to decline in the future and, as a result, any small risk associated with exposure to PAHs through consumption of subsistence resources near the spill zone should either remain constant or decline over time. The health hazards from PSP are much more serious than any health hazards associated with PAH exposure at the levels that were found. Two samples collected from Skan Bay contained PSP toxin at concentrations above the allowable level for commercial sale. Due to PSP concerns, the State of Alaska has implemented advisory warnings against the gathering and consumption of shellfish except at approved beaches since at least the mid-1990s. In mid-April 2006, the ADPH released the public health evaluation of the results of the 2005 subsistence seafood sampling program, which was posted on the M/V Selendang Ayu Unified Command website, along with a press release of the results and a fact sheet explaining the significance of the results (Mauseth et al. 2008). The results were presented to subsistence consumers in a video conference call in the spring of 2006. While there were no detailed subsistence harvest surveys conducted in association with the Selendang Ayu spill, a cooperative public health assessment of subsistence foods resulted in the clear communication to the subsistence consumers that there was no public health risk from eating subsistence foods because the PAHs in subsistence resources from Unalaska Island were not present at levels of health concern.

5.2.3 Effects on Food Quality and Safety Contamination of food resources was a key concern for traditional harvesters following EVOS. Although harvesters were advised that food was safe to eat if they could not smell oil in the food, this assurance was not sufficient and concerns about food safety remained even after studies demonstrated that oil contamination posed low health risks to subsistence users (Integral Consulting, 2006). Between 1989 and 1994, testing of subsistence foods showed no or very low concentrations of petroleum hydrocarbons (EVOSTC, 2010). The primary concerns about contamination directly after the spill were for marine mammals and shellfish and these fears had largely subsided by 2003. However, despite available scientific evidence on food safety, many subsistence users remain concerned about food contamination, especially clams, and even attributed the increase in paralytic shellfish poisoning to EVOS (Integral Consulting, 2006).

Page 5-70 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

5.2.4 Traditional and Cultural Activities Marine oil spills can have direct and indirect effects on the human environment including the harvest of wild country foods, fish, game, and botanical resources; the sharing of knowledge, ideas, and traditions among youth and elders; and local peoples’ relationships with the land and seascape, both psychologically and psychosocially. These potential effects can be complex and interrelated, making it difficult to quantify their extent and magnitude in terms of single drivers, dependent and independent variables, and cause and effect relationships. For example, the mortality of oiled wildlife has the direct effect of reducing the availability of those resources within the local food system, at least in the short term. Oil spills can affect multiple species simultaneously so it is possible that many wild food options within a local community’s portfolio can be affected. Depending on the nature and magnitude of a spill and the effectiveness of a spill response, opportunities for people to harvest, process and share wild foods within their cultural system can be reduced. Where subsistence activities are suspended due to environmental contamination or substantial ecosystem change, crucial opportunities for the cultural transmission of knowledge may be lost. However, these effects are temporary (although local ecological knowledge may also change, and change is the norm rather than the exception). When local people must procure more foods from the store to compensate for the loss of country foods, then the associated short- and long-term health effects of eating lower-quality industrially processed foods can affect health, tradition and culture. If the natural environment is affected to the point where the long-term ability of local people to conduct traditional harvest and associated cultural practices is compromised, changes to individual and/or community relationships with homelands, resources and the environment might result. However, existing information cited in this review does not indicate that marine oil spills pose long-term unrecoverable effects on either the natural or human environment. These various spill effects can create combined stresses on psychological and psychosocial health, but such stresses, however defined, are not inherently linked only to specific marine spill events. They may well also be affected by other resource development activities, the effects of social and cultural change, lack of employment opportunities in rural communities, or by severe storms, coastal erosion and community disruption and other catastrophic natural events. While it may seem like a simple logical leap to frame issues like alcoholism, depression, and domestic violence as being caused by a spill event, there are multiple drivers at work. Simplistically linking cause to effect in the context of traditional and cultural activities and cultural and social change, may well lead to spurious scientific conclusions.

5.2.4.1 Defining Traditional and Cultural Static definitions of concepts such as “traditional,” “customary,” and “cultural” must be avoided because of the danger of temporally fixing people’s lives and livelihoods to some imagined concept of tradition and indigenousness. It is inaccurate to ascribe traditional and cultural practices as belonging only to indigenous or aboriginal peoples. Likewise, it is inaccurate to think about tradition and culture as things that are only rooted in the past and that do not and/or should not change. Often for rural, indigenous or aboriginal livelihoods what is most traditional

July 2012 Page 5-71

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments is a culture of flexibility, innovation, and change. However, when specific technologies, techniques, or activities are labeled as “traditional,” they are effectively extracted from an evolving context of local lifeways and livelihoods, and are placed into an artificial category that is reified, by law, public opinion, or both, and justified through a perceived need for cultural conservation based on a static model of what culture is (Wooley 2002). Rarely do aboriginal residents divide their daily activities along lines that are clearly defined as modern or traditional, “for subsistence” activities or otherwise, though contemporary resource management regimes often require that they do so, with Alaska a classic case in point. Rather, they simply did or do what was and is necessary to make a living for themselves and their families, working on landscapes in and around their local communities (Loring and Gerlach 2010). For many, subsistence or country food harvesting as the preferred form of life integrates worldview, culture, and practice, a fact not widely appreciated by early and even modern Europeans who tend to view it simply in terms of technical skill. Thus, when discussing “traditional” and “cultural” activities in the context of a spill event, it is essential that the focus should be on maintaining connections and relationships with the land and seascape, and to protect and maintain patterns of flexibility, innovation, and learning, rather than simply conserving some externally-identified list of activities and technologies.

5.2.4.2 Nutrition Transition Indigenous peoples around the world are experiencing “the nutrition transition.” This concept refers to the progressive and systematic dietary and cultural changes away from foodways oriented to country foods and subsistence activities toward “western” diets based on industrially- produced store bought foods. This transition coincides with epidemic increases in biophysical and psychosocial ills including chronic obesity, coronary heart disease, type-II diabetes, depression, alcoholism, and domestic violence, which together have been termed the “new world syndrome.” This transition is something of a vicious cycle, whereby increased reliance on store- bought foods increasingly undermines participation in customary and traditional food harvesting and gathering activities, because of the time needed to earn wages to spend on store-bought foods. An oil spill events could contribute to this transition. Anything that undermines local people’s abilities to rely on locally caught fish and game accelerates the transition and further locks people in to the alternative cash economy. Davis’ (1996:256) description of weight gain by some local residents who had access to large amounts of freely distributed food during EVOS is an example. While a spill event will not likely be the sole or even primary driver of such a transition, its potential contribution to the process is clear, at least in the short run. Achieving recovery in the human environment could not possibly involve quantifying the proportion of the transition “caused” by the spill, and then somehow only compensating for that proportion of change, a fraction that is unknowable in even the best of circumstances. A more holistic, restorative justice approach argues that the best response would be for the responsible party to work together with the people affected to find solutions to the transition itself, as well as other human environmental

Page 5-72 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

effects. In this way, an integrated response becomes the vehicle for meaningful social engagement rather than “reparations” in the most basic sense. The well-integrated Unified Command for the Selendang Ayu spill, which included members of aboriginal communities in the spill response addressing unique local concerns about traditional sites, cultural respect and traditional food safety and quality (Morris 2005), provides a model that, if applied on future spills, could mitigate negative aspects of the food transition and enhance recovery in the human environment.

5.2.4.3 Effects and Recovery on Traditional and Cultural Activities Following the EVOS Davis (2006) noted that EVOS generated “considerable emotional energy; massive sums of money; a flurry of new jobs; a gaggle of regulations; and extensive, continuing litigation. But, because of legal constraints connected with class action lawsuits ….. few results of social science research on the recovery of communities have been published.” With that in mind, the community surveys that were undertaken to evaluate the long-term effects of EVOS on traditional and cultural practices, focused on four indicators—pedagogy and the cultural transmission of knowledge, learning, food and resource sharing, and local perceptions of impacts on tradition. The first indicator specifically refers to the extent to which the ability of elders to teach subsistence skills and values. In most communities, the respondents indicated that the influence of elders has remained stable, or has been declining most probably due to changes in demography (declining numbers of elders) or because of cultural changes in the overall community (Integral Consulting, 2006). However, in two communities, survey results suggest that, despite an initial decline in the influence of elders, the 2004 survey showed the majority of respondents perceived the influence of elders to have increased. A second indicator is the extent to which the spill affected the ability of young people to learn traditional culture. The survey results showed that, shortly after the spill, only 39% of respondents in 1992 stated that youth were learning the necessary skills, but this percentage increased to 48% in 1998 and 53% in 2004 (Integral Consulting, 2006). The reasons why youth were not learning the necessary skills related to lack of youth interest, a lack of teachers and cultural changes in the overall community. The third indictor is whether or not subsistence food sharing patterns in the communities changed as a result of the spill. The best available information suggests that when harvest levels declined as a result of the spill, vulnerable members of the community (elders, single mothers with dependent children and inactive single persons) were given priority during food sharing. Surveys conducted in 1998 concluded that 28% of respondents thought that food sharing had declined, 43% thought it had remained the same, and 24% thought that it had increased. By 2003, food sharing had become more widespread; with 70% of respondents in 10 communities reporting that food sharing had stayed the same or increased (Integral Consulting, 2006).

July 2012 Page 5-73

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

The fourth indicator is whether survey respondents believed that the spill had adversely affected their traditional way of life. In 2004, 83% of respondents said the traditional way of life had been injured by EVOS, 74% percent reported that recovery of the traditional way of life had not occurred. While the results of the surveys suggest that EVOS may have adversely affected traditional and cultural activities and continued to do so as recently as 2004, the reported effects on the traditional way of life are somewhat inconsistent with survey responses for the other three indicators. It has been suggested that this may reflect the inability of survey respondents to differentiate the effects of EVOS from other factors that have been affecting their way of life (Integral Consulting, 2006). These factors include changing demographics, increased competition from recreational and commercial users for available resources, and continued concerns about food safety, with all of these having a potentially wide range of causes. Other reasons may be related to the confounding issues that Davis (1996:233) noted. An alternative explanation is that it is both impossible and undesirable to try to tease out the effects of an environmental on local peoples and cultures from the broader context of socioeconomic challenges and trends. One cannot legitimately speculate on how social, cultural, and economic circumstances would have continued to change for a community had a spill not occurred, in part because one cannot anticipate surprise and/or innovation. In other words, it is impossible to know how community members may have worked to improve local circumstances had a spill never occurred. Spill events must therefore be considered as events that bisect a timeline and initiate a new path in history. Thus, rather than becoming mired in debates about causality (the pursuit of which is presumably driven by a criminal model of justice preoccupied with punishing the offender for their specific transgressions), a more positive approach is to consider cultural and traditional practices from the perspective of restorative justice, a perspective that focuses on ensuring the human environment is on a path toward healing and recovery. Integrating members of aboriginal communities in the ongoing process of spill response planning and enhancing local spill response training strategies provides an important aspect of this process, and one that has become standard in Alaska.

5.2.4.4 Summary on Recovery of Traditional and Cultural Activities These findings represent the most recent conclusions upon which the EVOSTC has based its assessment of the recovery of traditional and cultural activities. Since the spill, numerous authors have posed separate theories and models about how Aboriginal communities have been affected by technological , such as EVOS. A central theme in the literature (see Impact Assessment Inc., 2001) relates to the “alienation” of Aboriginal people from their traditional and culture activities as a result of such an event. This alienation can be caused from a variety of factors: • The disruption of subsistence harvest practices threatens the ability to transmit culture and tradition to younger generations and the event both symbolizes past threats to Aboriginal culture and raises questions about the future ability to maintain their identity.

Page 5-74 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

• The change in environmental conditions as a result of such an event lies outside the range of traditional knowledge, experience and tradition, and the advice from outside experts can call this traditional knowledge into question and also provide information that is sometimes contradictory and inconsistent with common sense. • Loss of particular subsistence foods that are of cultural importance may result in the consumption of replacement food with no cultural importance or subsistence foods that may be or perceived to be contaminated. • Concerns about the long-term effects of consuming contaminated foods. Based on the review of the effects of EVOS, Impact Assessment Inc. (2001) concluded that: “...Natives experienced the disruption of subsistence as a threat to their ways of life. The effect was to alienate individuals from the cultural identity from key social practices and interactions, spiritual and other values about natural resources, and traditional knowledge about the biophysical environment.” They further note that a return to pre-spill harvest levels does not negate the damage on culture that has occurred and that EVOS had adverse consequences for individual and communal life.

5.2.4.5 Recreation and Tourism The environmental damages resulting from EVOS caused resource managers to limit recreational hunting and fishing activities in specific areas and some marine users (boaters and kayakers) were precluded from some areas. There was increased congestion in those areas that were still available for recreation and tourism (EVOSTC, 2010). Following EVOS, various studies were commissioned to assess the potential effects on recreation and tourism in Prince William Sound. An Alaskan survey of sport fishermen (Mills, 1992) concluded that, for the five years prior to the spill (1984 to 1988), the number of anglers in the area affected by the spill was increasing at 10% per year, the number of days fished was increasing by 8% per year, and fish harvests were increasing at 14% per year. In 1989, the year of the spill, the number of anglers decreased by 13%, the number of days fished declined by 10% and the fish harvest dropped by 10%. The results showed that, while the amount of angling by resident fishermen decreased in 1989, the number of non-resident anglers actually increased slightly. It is noteworthy that 10% of the people who fished in the area during 1989 were oil spill workers. Sport fishing data for 1990s indicate that there was an increase in sport fishing activities, but the increase was not as large as would have been expected in the absence of EVOS. A related study was undertaken by Carson and Hanemann (1992) who attempted put a dollar value on the loss of recreational fishing related to EVOS. They ultimately estimated the losses at between $3.6 and $50.5 million during 1989 and 1990, based on the reduction in the number of days of recreation and assuming an average value of $250 per day. In arriving at this range of values, they note that the reduction in angling activity by resident anglers could be related to a variety of reasons, including a reduction in the quality of recreation resources, boats being diverted for clean-up purposes, increased congestion at areas that remained available for fishing,

July 2012 Page 5-75

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments and lack of time because of participation in clean-up activities. For non-resident anglers, the increase in activity was thought to be related to the effects of advertising prior to the spill, fishing trips having been pre-paid, and the influx of people for clean-up activities. A third study undertaken by the McDowell Group (1990) examined the effects of the spill on tourism. The study concluded that EVOS had both negative and positive effects on tourism. Spill clean-up activities resulted in demands for labour, accommodation, vehicle rentals, air taxis and boat charters, and this provided a source of income for some businesses in the region. However, 43% of businesses (lodges and resorts, tour companies and guided outdoor activities) were either located away from the site or were unable to provide clean-up related services, and reported that they had been substantially or completely affected by the spill. Due to lack of accommodations, charter boats and air taxis for visitors, 59% of businesses reported spill-related cancellations and 16% reported less business than usual. There was also a severe labour shortage as people were hired to undertake clean-up work, resulting in higher labour costs for tourism operators. It is estimated that the spill resulted in a loss of 9,400 visitors in the summer of 1989 with a loss of $5.5 million in visitor spending. For the summer of 1990, a year after the spill, the adverse effects of the spill were less severe, although bookings remained lower than prior to the spill and 12% of businesses indicating that they were still significantly of completely affected by the spill. According to (EVOSTC, 2010), recreation and tourism in the affected area will have “recovered, in large part, when the fish and wildlife resources on which they depend have recovered, and recreation use of oiled beaches is no longer impaired”. Although more than $10 million was spent by 2001 on repair and restoration of recreational facilities in the spill area, and damage caused by the spill or clean-up efforts, EVOSTC concludes that, as of 2010, recreation and tourism is still recovering from the spill but had not yet recovered. The main reason for this conclusion is that some fish and wildlife species of importance to tourism (harbor seals, Kittlitz’s and marbled murrelet, pigeon guillemot, clams, mussels, harlequin ducks, sea otters and killer whales) have not yet recovered and some beaches used for recreation still contain residual oil. While there are still some residual effects of the spill affecting tourism and recreation resources, the levels of activity are increasing. Integral Consulting (2006) reported that recreation and tourism participation in the spill affected region was high in 2003 and had surpassed pre-spill levels. The amount of marine recreational fishing in Prince William Sound increased from 30,383 angler-days in 1978 to 173,554 in 2004, suggesting the sport fishery has recovered (Figure 5.12). Levels of non-consumptive recreation and tourism have continued to increase in the area affected by the spill area, despite a decline in wildlife sightings. The amount of wildlife viewing, scenic driving, off-road driving, and biking in the Chugach National Forest area was increasing beyond pre-spill levels at a rate slightly greater than the population growth rate. As a result, Integra Consulting (2006) observed that available information suggests that Prince William Sound is becoming increasingly popular for recreation and tourism and that lingering oil does not appear to deter visitors.

Page 5-76 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Source: Integral Consulting (2006) Figure 5.12 Estimated Fishing Effort for Prince William Region from 1977 to 2004 There appears to have been no quantification of the extent to which the Selendang Ayu spill affected recreation and tourism in the surrounding area. An assessment of potential losses (Kohout and Meade 2008) observed that recreational activities by residents of the area was impaired by response activities, fisheries closures, Coast Guard access restrictions and concerns about oiling. Regarding tourism, it was concluded that, while some tourists may change their plans to avoid the area, they might choose to visit other parts of Alaska that offer comparable recreation opportunities (Kohout and Meade 2008). Gill and Ritchie (2006) reported that there was decline in tourism as a result of the Selendang Ayu spill and described 2005, the year of the spill, as being a “lost season”. However, it was expected that most ecotourism businesses would resume activities in 2006 and that ultimate recovery from the spill will depend on the extent to which the tourism client base was affected by the spill. Overall, available information from the two Alaska spills indicates that recreational and tourism activities do recover from marine oil spill events, with the duration of the recovery being dependent on the length of time required for fish and wildlife populations to recover, as well the time required for evidence of oil on beaches to disappear. This conclusion is consistent with information from other marine spill events. For example, information for the 2010 BP Deepwater Horizon spill in the Gulf of indicates that tourism expenditures in were expected to return to normal conditions by early 2013, with the total value of leisure tourism losses estimated to be $691 million, although this will be partially offset by a $395 million increase in expenditures by business-based tourism related to oil spill clean-up (Anderson, 2011). Another report suggested that, while tourists avoided the affected areas during 2010, tourism numbers in 2011 were reported to be only 20% below 2009 levels (Amy, 2011) and that tourism expenditures were again rising due to a general improvement in economic conditions in the region and people wanting to return to the beaches after having been forced to avoid the area during the previous summer. In affected parts of

July 2012 Page 5-77

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Florida, reports indicate that levels of tourism were back to normal one year after the spill, due in part to advertising campaigns partially funded by grants from BP (USA Today, 2011). In , it was reported that 150 to 350 tonnes of oil spilled from the container ship Rena in October 2011 was unlikely to affect the country’s tourism industry, although there may localized effects (English.news.cn, 2011). The spill was not expected to pose a threat to New Zealand’s “clean, green image” and, with appropriate marketing, could reinforce the images of its beautiful environment because, as was the case with EVOS, news coverage showed people both the effects of the spill and the beauty of wilderness areas. A study of the combined effects of oil spilled from the Ixtoc 1 drilling platform (454,000 tons) and from the sinking of the Burma Agate (8,440 tons) in the in 1979 concluded that tourism losses amounted to about $4 million while recreation losses totalled $3 million, with most of the losses being experienced by businesses located close to the water (Restrepo & Associates 1982). The report also suggests that extensive media coverage may have also resulted in tourism losses. In reviewing the literature, one of the challenges in assessing the potential effects of a marine oil spill on recreation and tourism is the lack of pre-spill information on extent and location of use. While governments typically report commercial and recreational fishing information at a regional or sub-regional basis, detailed regional information on the nature and extent of other recreation and tourism activities is usually not available. This means that researchers often have to try to reconstruct pre-spill conditions from available information after a spill has actually occurred, and this can prove very difficult in remote areas.

5.2.5 Passive Use Values It is argued that an oil spill can have effects on people that have not, do not and may not actually make use of areas or resources affected by spills. These effects are termed “passive values” and are described as the “appreciation of the aesthetic and intrinsic values of undisturbed areas and the value derived from simply knowing that a resource exists” (EVOSTC, 2010). In the case of an oil spill, the effects on passive values can be measured in terms of the public perception of the damages to the aesthetic and intrinsic values of the affected area. Recovery following a spill was defined by EVOSTC as being when “people perceive that aesthetic and intrinsic values associated with the spill are no longer diminished by the oil spill”. As of 2010, the EVOSTC still considers that services related to passive use have not recovered because recovery of a number of injured resources (fish and wildlife) has not yet occurred. While it is conceivable that an oil spill could damage passive values, it is very difficult to measure these values. A study of the potential effects of EVOS on passive use values was undertaken for the Attorney General of the State of Alaska in 1992 (Carson et al, 1992) and concluded that damages associated with the spill amounted to $2.8 billion. This estimate was based on the results of a contingent valuation survey that asked households in the United States what they would be willing to pay to implement a spill prevention plan that would minimize the possibility of a catastrophe similar to EVOS from ever happening again. While the EVOSTC has undertaken a variety of efforts to restore passive values, there have been no subsequent studies or

Page 5-78 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments surveys to assess the extent to which passive use values have recovered (Integral Consulting, 2006). While the contingent valuation approach used by Carson et al. (1992) to measure passive use values was, at the time, considered ground-breaking in many ways, this approach was also known to have a variety of methodological problems and other approaches based on choice experiments and stated and revealed preferences have since evolved. Some of the flaws in the original study were identified by Carson et al. (2004) in a subsequent assessment of the potential effects of hypothetical oil spill off the coast of California. A comparison of the two approaches shows that passive value estimates derived from one study cannot be applied to another situation because, among other reasons, the economic values resulting from different spill scenarios at different locations involving different populations are fundamentally not comparable. Another observation is that many of the elements of the hypothetical spill prevention plan used in the 1992 study to assess willingness to pay have since become part of standard operating procedures for marine tankers. Thus, it is evident that the reported estimates of passive value damages are unique to EVOS and dated, and cannot be applied in other situations. A more detailed discussion about the measurement of passive use values associated with oil spills in the context of the Northern Gateway project are provided by Wright Mansell Research Ltd. (2012).

5.2.6 Archaeological Resources Marine oil spills and the resulting clean-up activities can adversely affect archaeological resources. In the case of EVOS, the oil spill region (including oiled and unoiled areas) was known to contain more than 3,000 site of archaeological or historical importance and 24 of these sites were known to have been affected in some way by spill clean-up activities or looting and vandalism related to the clean-up (EVOSTC, 2010). A Cultural Resource Program including input from tribes, the Alaska State Historic Preservation Office (SHPO) and agency archaeologists surveyed over 4,000 km of potentially oiled shorelines during the 1989–1991 field seasons (Haggarty et al. 1991). Hundreds of specific clean-up areas where there were cultural resources present were inspected and monitored. Archaeologists revisited two hundred previously known sites near oiled shorelines, and they located, documented, and reported 326 new sites to the Alaska Heritage Resource Survey increasing the number of known sites in the spill area (near oiled shorelines) from 283 to 609 (Haggarty et al. 1991). As disturbed archaeological resources cannot be replaced or restored, recovery was defined in terms of when the spill related injury has ended, looting and vandalism have ceased, and the remaining artifacts at vandalized sites have been preserved. Based on this definition, EVOSTC considered the archaeological resources affected by EVOS to have recovered as of 2002 (EVOSTV, 2002). The archaeological information collected and synthesized by the Program enabled state and federal land managers with cultural resource site-management responsibilities to both assess the damage that occurred, and to begin more robust research and management initiatives. The new procedures developed during the spill response were applied during subsequent spill responses.

July 2012 Page 5-79

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Many of the Cultural Resource Program policies and procedures were formally adopted in 2002 as the Alaska Implementation Guidelines for Federal On-Scene Coordinators for the Protection of Historic Properties during Emergency Response under the National Oil and Hazardous Substance Pollution Contingency Plan (Alaska Regional Response Team, 2002). These guidelines were in place during the Selendang Ayu spill response. In the case of the Selendang Ayu spill, archaeologists were employed on the response and clean- up effort on behalf of the responsible party, the US Coast Guard and local tribal authorities. They worked with the SHPO and the local tribal members to ensure that any activities conducted during this clean-up did not disturb known historic sites. They also documented a number of new sites (Morris 2005). Archaeologists reported that there was no effect on sites from the oil or clean-up activities (Kohout and Meade 2008).

5.2.7 Effects on Industrial or Other Land and Resource Uses Aside from effects on commercial fisheries and tourism, EVOSTC does not describe the effects of EVOS on other industrial or land and resources uses. However, one of the obvious effects of the spill, containment and clean-up was a short-term disruption in tanker traffic moving in and out of Valdez. Tanker traffic resumed five days after the spill occurred, but only during daylight hours with a two-tug escort and avoiding clean-up activities by at least 500 yards (Kelly 2002). For the Selendang Ayu spill, the Coast Guard initially projected that oil would not affect the fish processing industry in Unalaska. However, some tar balls did enter Unalaska Bay, forcing processing plants and crabbing vessels to monitor their water intakes and remove contaminated materials. As a result, this industry was able to continue operating during January’s critical crab season).

5.2.8 Social Effects Oil spills represent a type of what some academics have termed “technological disasters” and various models have been developed to predict the effects of such disasters on communities. Gill and Ritchie (2006) applied one of these models to assess the community responses to the Selendang Ayu spill. They note that seven factors characterize community responses to technological disasters like spills, including disruptions, emotional responses and erosion of what they term “social capital.” When Gill and Ritchie (2006) applied this model, based on information taken from interviews, observations and media accounts, they found no evidence of six of the seven types of effects predicted by their conceptual model. They conclude that the Selendang Ayu spill did not result in collective trauma and stress, a corrosive community, negative lifescape change, secondary trauma, or loss of social capital in affected communities. Consequently, they suggest that there is little to no evidence suggesting that the Selendang Ayu spill will have any negative long-term social effects and there may actually been benefits, in terms of increased appreciation for the communities ties to the natural environment. It was suggested that the community of Unalaska was highly resilient to the effects of the Selendang Ayu spill because of its previous history of

Page 5-80 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

and exposure to natural and other technological disasters. Davis (1996, 2006) made similar observations regarding Alutiiq region villages.

5.2.9 Effects of Clean-up on Human Environments For marine spills, the indirect effects of clean-up operations on the human environment can be more profound than the direct effects of the spill. The effects on the human environment, and the resulting recovery time, can be very pronounced when a large clean-up effort is launched in and among a number of small relatively isolated communities. While the sudden new demand for labour, goods and services can benefit communities, the resulting refocusing of local economic activities and the introduction of large numbers of non-local workers can cause considerable social disruption. The clean-up activities associated with EVOS provide numerous examples of benefits and costs to the human environment. Overall, clean-up activities commenced in April 1989 with the last major shoreline clean-up efforts occurring in 1992, three years after the spill (Impact Assessment Inc. 2001). The total cost of the clean-up totalled $2.5 billion, with $2.0 billion being spent in the first year (ITOPF 2011b). At the height of operations, more than 10,000 workers were employed in response operations associated with the spill (ITOPF 2011b). While response operations had substantial effects on the human environment, these effects were magnified by the way in which clean-up operations were managed. Under legislation in effect at the time of the spill2, managing and coordinating an oil spill in Alaska coastal waters was the responsibility of the U.S Coast Guard, and the contingency plan for dealing with an oil spill had been developed between the State of Alaska and the Alyeska Pipeline Service Company (Piper and ADEC, 1993). However, Exxon, as one of Alyeska’s parent companies, assumed control of clean-up on the second day of operations and chose not to implement the approved contingency plan, leading to considerable confusion among responders and within the affected public, and resulting in a “confused and ineffective response” (Piper and ADEC 1993). Ultimately, “the Coast Guard and Exxon agreed on a management system that provided for more federal involvement and direction, but allowed Exxon to write the checks” (Piper and ADEC 1993). Under this arrangement, the governments agreed to relinquish some of their authority and allow Exxon to use its financial resources to purchase and transport equipment, hire vessels, and deploy additional clean-up contractors. This arrangement was more effective in terms of obtaining the extensive resources immediately needed for the response but led many of the affected parties to distrust Exxon, which had caused the spill in the first place, and expect the state and federal governments to provide support and assistance which never came (Piper and ADEC, 1993). These circumstances are likely to have contributed to the public angst about the spill and provided additional motives for various aggrieved members of the public to later seek punitive damages.

2 The National Contingency Plan was established as part of the U.S. Clean Water Act of 1973 and established as series of regional authorities to oversee operations in principal zones of marine traffic.

July 2012 Page 5-81

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

A second problem with the initial spill response was that there were insufficient resources to contain or clean-up a spill of the magnitude of EVOS. This was evident at the time of the event and eventually formed the basis for a successful lawsuit against the Alyeska Pipeline Service Company for loss of Alaska tax and royalties because of the interruption of tanker traffic and for having insufficient spill-fighting equipment (Los Angeles Times, 1992). Two days after the spill, the Alaska Department of Environmental Conservation initiated its own clean-up efforts by including the local fishing community in efforts to prevent the oil from reaching fish hatcheries in Sawmill Bay. These activities included more than 50 Cordova fishermen, the village of Chenega Bay, the Prince William Sound Aquaculture Association, 60 Alaska Department of Conservation and contract staff, 40 private vessels, two state ferries, the Alaska National Guard and the Alaska State Troopers (Piper and ADEC, 1993). About 18 days after the spill, Exxon submitted a preliminary shoreline clean-up plan that called for using up to 4,000 workers in land-based camps. This plan was eventually changed, with clean-up workers being housed in vessels, barges and other watercraft, to prevent disruption of archaeological sites, encounters with wildlife, waste management issues, and disagreements over land use (Piper and ADEC, 1993). Exxon hired individuals from affected communities to work on the response, and priority was given to Alaska Natives, commercial fishermen, and other affected Alaskans. In 1989, 43% of all adults from the 15 Alaska Native villages in the spill area were employed in oil spill clean-up and these activities generated 50% of annual income for these villages (Fall et al. 2001). While the response provided an alternate source of employment for some people who would otherwise have been unemployed because of oil-spill related fisheries closures, it created numerous other problems. The opportunities for relatively high paying clean-up jobs meant that more adults in Cordova were working, leading to increasing demands for child care, but the child care workers had also quit to work on the spill (Piper and ADEC, 1993). Similar labour shortages and the rising labour costs were experienced by other businesses in the area, notably tourism (McDowell Group, 1990). However, not everyone in the local communities who was adversely affected by the spill was able to participate in clean-up efforts. Piper and ADEC (1993) note that, although skiff owners could earn $20,000 during the clean-up season and large vessel owners could earn “10 times that”, local fleets struggled to divide the work up among available vessels and the majority of the 300 vessels based in Kodiak were not involved in the clean-up because very little oil made it that far. Some communities reported animosity between fishermen who were hired for clean-up activities and those who were not, and some crewmen were treated differently from boat owners (Impact Assessment Inc., 2001). In addition, many Alaska Natives felt they did not benefit directly from the response, having received very little in terms of clean-up work or compensation, even though millions of dollars were being spent to rescue and rehabilitate wildlife (Piper and ADEC, 1993) The high wages and demand for labour resulted in attracting people from other Alaskan communities and other states. It is estimated that the population of Valdez doubled, with workers, bureaucrats, oil executives and the media renting all the rooms in town, resulting in

Page 5-82 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments local residents hiring out beds and rooms (Pulliam, 1989). Some local residents were displaced from their normal housing because their landlords were able to find imported workers who were willing to pay more (Phillips, 1999). The new workers increased business for restaurants and bars, but there was an associated increase in bar fights, and thefts (Impact Assessment Inc., 2001; Pulliam, 1989). The demand for food and housing was so great that the state provided meals and housing (both transient and seasonal) for its workers in Valdez (Piper and ADEC, 1993). The economic effects of the spill were considered a “boom” for local businesses, with increased demand for consumer goods, hardware stores and sporting goods stores. Clean-up activities also benefitted the State of Alaska, with a drop in the unemployment rate and the associated revenues providing a brief economic respite from several years of economic stagnation resulting from low world oil prices (Piper and ADEC, 1993). With the increased temporary population came an increased demand for municipal services, resulting in a major increase in the workloads for the police, fire department, parks managers, garbage collectors, road and street maintenance workers, dock managers and harbour masters (Piper and ADEC, 1993; Impact Assessment Inc., 2001). City workers were also expected to serve on spill coordination activities. At the same time, the costs of providing these services increased, so municipal governments were forced to seek repayment for costs from Exxon and the Alaska government. Initially, there was no standard or equitable system for various municipalities to easily obtain repayment of expenses from Exxon, with the larger, better organized communities being able to obtain compensation more quickly than the smaller communities (Piper and ADEC, 1993). This inequity resulted in the mayors of the communities developing a standard community agreement that was never accepted by Exxon. As a result, municipal governments sought financial help from the state, but the monies received could only be used to fund activities related to the spill response and not for activities related to spill-related demands for services (Piper and ADEC, 1993). Thus, the spill response placed major demands on municipal governments and elected officials, and these effects lasted beyond the duration of clean-up activities. There were other social issues related to the clean-up. Residents of some communities, notably Valdez, noted that their communities had been invaded by outsiders, generating concerns about public health and safety, and that they had “lost control” of their daily lives (Impact Assessment Inc., 2010). Adding to community stress was the “pervasive presence of Exxon ‘security guards’” and restrictions placed on resident activities in their home communities, as well as Exxon’s perceived “heavy handedness” in its interactions with residents, businesses and local governments (Impact Assessment Inc., 2010). The overall community response to EVOS clean- up activities was described as: The stress levels, the sudden flow of money, the influx of transients with no obligations or connections to the communities, all helped make day-to-day life in the spill towns in 1989 a bizarre, unsettling, and occasionally dangerous experience. (Piper and ADEC, 1993)

July 2012 Page 5-83

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

In summarizing the effects of clean-up on affected communities, Impact Assessment Inc. (2010) made some general observations: • The adverse effects associated with the spill and clean-up activities were lower in communities that had more diversified economies. • The benefits and costs of spill clean-up were not equally distributed among businesses within communities. • Smaller communities and smaller businesses were more vulnerable to economic losses. • Economic losses to commercial fishermen were not evenly distributed because not all vessels and their crews were engaged in clean-up activities. • The spending in communities associated with clean-up activities created a variety of social impacts and did not mitigate all the losses. • Local governments experienced adverse fiscal impacts. Overall, the evidence associated with EVOS suggests that the indirect effects on the human environment associated with the spill response can have just as great an effect as the direct effects of the spill itself, although the response effects tend to be of shorter duration. But, to some degree, the human environmental effects associated with EVOS spill clean-up were unique to that unprecedented situation, because of the size of the spill and the unique way in which clean-up operations were conducted. Although the Selendang Ayu spill occurred 15 years after EVOS, some of the human environmental effects of the response on the human environment were similar – although on a relatively small scale. The observed effects included (Impact Assessment Inc., 2011): • Unequal distribution of clean-up benefits and costs among commercial fishermen. Only some captains were qualified for clean-up activities and they prospered while others were not able to participate. • The introduction of clean-up workers into the community resulted in some temporary crowding, limited availability of goods and services, and changes from the normal duties and ways of life for some local residents. • Increased demands on local governments and public sector services. • Resistance to external authority by local people to newly arrived response officials. In the case of the Selendang Ayu spill, these effects were generally short-term, lasting at most two years after the spill, although some personal rifts continued. In considering why these effects still occurred, despite the lessons learned from EVOS, Impact Assessment Inc. (2011) observed that, while having detailed contingency plans may ultimately lessen the effects on the human environment if a spill actually does occur, it is not possible to adequately prepare a community for such an unknown event at an unknown place and time. Impact Assessment Inc. (2011) also concluded that understanding of the human effects resulting from spill events and associated clean-up activities must be incorporated into emergency response plans. They note “social

Page 5-84 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

problems occurring during the early phases of oil spills can persist and alter the trajectory of social and sociopolitical relationships in the future” and conclude that the importance of including the human element into response planning “cannot be overstated”. In mitigating these potential human environmental effects, Impact Assessment Inc. (2011) also noted that an effective Unified Command coordinating response work among many agencies and entities is essential. The Selendang Ayu accident and spill revealed an initial period of tensions between visiting responders and local residents, but that local perspectives changed over time with local people reporting their belief that the constituents of the centralized command process were working in the interest of the community with a high level of dedication of the response team.

5.2.10 Effects of Compensation and Litigation The last stage in the recovery of the human environment following a marine oil spill relates to the time required to settle all outstanding claims for compensation from damages. The length of time required to reach a final settlement depends on how quickly the parties can come to an agreement on compensation for damages. Resolution of outstanding issues can involve a considerable amount of time if the parties ultimately rely on the courts to determine an appropriate settlement. This was particularly true for EVOS. While Exxon began paying compensation to individuals and companies for damages shortly after the spill occurred, some of the claims still had not been resolved 20 years later. These delays in reaching final settlements can largely be attributed to the laws in place at the time EVOS occurred and the US justice system. To understand the effects that payment of compensation and the associated litigation related to EVOS had upon the human environment, it is necessary to describe the each element of the compensation and litigation process.

5.2.10.1 Compensation for Direct Damages According to Impact Assessment Inc. (2010), Exxon paid $300 million in compensation to more than 11,000 people and businesses in the spill-affected areas within one year of the spill. Some payments were made as part of compensation programs implemented by Exxon to address communities and individuals that were directly affected by the spill. Others payments were made in response to the 150 law suits initiated by various affected parties within eight months of the spill that were settled out of court or adjudicated (Impact Assessment Inc., 2010). There were some issues with Exxon’s compensation program, however, especially for commercial fishermen. One of the first issues was that it was not clear whether fishermen who chose to participate in and get paid for the clean-up would be eligible for compensation, so some fishermen chose not to participate in the clean-up. In addition, Exxon was only prepared to provide compensation to fishermen who were ready to fish when the closures occurred. So, in anticipation of future closures and potential compensation payments, some fishermen chose to be ready to fish (taking on the costs of provisions and fuel) with the hope that they would receive compensation when the closure was announced. The net effect was that fishermen who

July 2012 Page 5-85

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments participated in the clean-up made more money than they would have through fishing, and fishermen who did not work in the clean-up and hoped to receive compensation, earned less (Piper and ADEC, 1993). This led to some long standing issues among individuals within the commercial fishing industry. A second problem was that some fishermen had difficulty proving claims for compensation because they lacked records showing that they had worked in previous years. New entrants into the fishery were unable to provide grounds for compensation, as were people who had been working but had not been paying taxes. Furthermore, payments were based on previous harvest levels and had no bearing on what they could have caught in 1989 or on the skill level of the particular fishermen (Piper and ADEC, 1993). Furthermore, some crewmen, boat owners without permits, cannery workers and small businesses dependent on commercial fishing were not eligible for compensation because they could not produce the fishing permits upon which Exxon based its compensation (Alaska Daily News, 1989). A third issue was that the compensation could not clearly account for the changes in cash flow and debt servicing requirements for individual operators, and this resulted in financial hardship for some operators. Initially, Exxon provided compensation to boat owners, with the expectation that they would share the settlement with their crew. This did not always happen so Exxon modified the program so that skippers would have to provide a list of eligible crew members and Exxon would pay them directly. This resulted in some boat captains inflating their crew lists with family members to maximize their payments (Alaska Daily News, 1989). Although Exxon was aware of these problems with the compensation system, it did not see payment of fair and equitable compensation within the commercial fishing industry as being its problem (Alaska Daily News, 1989). Despite the efforts made to compensate individuals and companies affected by the spill, not everyone was happy with the outcome. Some people received no compensation whatsoever while others were dissatisfied with the amounts they received. As a result, thousands of people attempted to seek compensation by seeking punitive damages from Exxon through the courts. These efforts are described below.

5.2.10.2 Criminal Liability One mechanism for seeking restitution for damages from EVOS was to charge Exxon with various criminal charges. In 1990, Exxon and its shipping subsidiary were indicted on five criminal violations. These included two felony charges related to allowing an incompetent crew to leave port, and three misdemeanours, related to killing waterfowl under the federal Migratory Bird Treaty Act and discharging oil under the Clean Water Act and Refuse Act (, 2010). Although Exxon pled not guilty, it reached a plea bargain in 1991 that resulted in fines and restitution valued at $250 million, although $125 million was forgiven based on Exxon’s response efforts. The resulting payment of $125 million included $50 million in punitive fines and $100 million in restitution to the state and federal governments to cover their clean-up costs (Impact Assessment Inc., 2010).

Page 5-86 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

5.2.10.3 Civil Liability The United States and the State of Alaska also filed court actions against Exxon and Exxon Pipeline to recover damages for injury to natural resources arising from the oil spill according to provisions under the Clean Water Act. A settlement agreement among the parties was reached in 1991 and it saw Exxon agree to pay $900 million in natural resource damages. These payments were to be paid in equal installments over 10 years, with the last payment being made in September 2001 (EVOSTC, 2011). The payments from Exxon were put into a trust fund, with $684.1 million being made available to the EVOS Trustee Council for restoration efforts to be undertaken after 1992 and $215 million applied to clean-up costs borne by government agencies. One of the ways in which the EVOSTC could use the fund was for “acquisition of equivalent resources”, which could include everything from purchases of private lands to land management agreements with private parties or other government agencies (Piper and ADEC, 1993). Some 378,100 acres were eventually purchased from six Alaska Native corporations and entities on Kodiak Island for $239 million (Impact Assessment Inc., 2010). These purchases were not without controversy, however. There was disagreement within the corporations as to how the money should be spent and, when revenues were provided directly to individuals, this resulted in increased mobility of community members, a short-term reduction in employment in commercial fishing, and increased investment and participation in subsistence hunting and fishing (Impact Assessment Inc., 2010; Alaska Daily News, 1999a). Thus, even the payment of funds to support environmental restoration can have unintended consequences for the human environment. The settlement agreement also contained a "reopener window" that allowed the government to make a claim for up to an additional $100 million to restore resources that suffered substantial losses or declines that were not known or anticipated at the time of the original settlement. The agreement contained a four-year window after the final payment for the governments to submit their additional claim and, on June 1, 2006, the U.S. and Alaska governments provided ExxonMobil Corporation with a detailed project plan for the clean-up of lingering oil at an estimated cost of $92 million (EVOSTC, 2011). Exxon was allowed 90 days to pay or respond. A 2009 news article (Environment News Services, 2009) indicates that, on the 20th anniversary of the spill and three years after the additional claim was submitted, the status of the claim was still not resolved and neither the US President nor the Governor of Alaska had taken any action to collect. There is no more current information on the status of the reopener on the EVOSTC website. A second civil suit was filed against the Alyeska Pipeline Service Company, which operates the Port of Valdez, by the U.S Government and the State of Alaska. As described earlier, the suit was for loss of Alaska tax and royalties because of the interruption of tanker traffic and for having insufficient spill-fighting equipment (Los Angeles Times, 1992). This case was settled in 1992, with $29.7 million being paid to the State of Alaska and $2 million to the U.S. Government (Impact Associates Inc., 2010).

July 2012 Page 5-87

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

5.2.10.4 Punitive Damages What makes EVOS and the recovery from the spill unique was the contentious litigation process resulting from law suits for punitive damages. Class action suits for punitive damages on behalf of 32,000 plaintiffs were filed with the federal court and trials commenced in 1994, five years after the spill. In 1996, the Federal Court jury in Anchorage awarded the claimants $5.2 billion in damages. This amount was contested by Exxon and, in 2001, the Ninth Circuit Court of Appeals reduced the amount to $4 billion. This amount was again appealed to the Ninth Circuit which, based on new punitive damage guidelines, decided in 2004 to adjust the amount to $4.5 billion in damages plus $2.25 billion in interest. Exxon again appealed, using the argument that, based on precedents, the ratio of punitive damages to actual harm (which was determined to be about $513 million based on the earlier compensatory damages verdict and pre-trial settlement) was excessive. As a result, the punitive damages settlement was reduced to $2.5 billion, including interest. Exxon petitioned for a rehearing in 2007 and ultimately, in 2008, the U.S. Supreme court ruled that punitive damages could not exceed compensatory damages, so the award was reduced to $507.5 million. In 2009, it was further determined that the plaintiffs should also receive interest. It was expected that actual payment of the settlement and interest would continue into 2011 (Impact Assessment Inc., 2010). This protracted process for dealing with punitive damages under the American legal system has had a significant effect on the human environment. Determination of the final settlement occurred 19 years after the spill occurred, the amount being paid amounted to 10% of what was initially awarded, and the settlement awards are subject to taxation. By the time the final settlement was reached, it is estimated that 20% of the original plaintiffs were deceased. In 1999, 10 years after the spill, news reports described some of the effects on the human environment that were occurring as a result of the protracted litigation process. The Alaska Daily News (1999b) reported that, shortly after the spill, lawyers were actively seeking clients, promising them large damage settlements. The large amounts of money being spent by Exxon to hire local residents for clean-up operations raised expectations for compensation, and the initial award of $5 billion would have seen many of the plaintiffs being paid $1 million or more in damages. Based on the expected compensation, some plaintiffs spent heavily on new boats, equipment, consumer items and holidays and other items that they could not pay for when the expected settlement did not occur and the revenues from commercial fishing declined. It was reported that, by 1999, 50 plaintiffs had filed for bankruptcy and 4,200 liens had been filed by creditors. As a result, there was widespread resentment of Exxon, for having caused the spill and drawing out the settlement process, and for the legal system for allowing the process to be drawn out. There was also resentment and anger within the communities because some of the fishermen who were well-paid for their participation in clean-up activities (sometimes referred to as “spillionaires”) were financially stable enough to survive the wait for punitive damages while fishermen who expected compensation went bankrupt, had to take a second job, or chose to move away.

Page 5-88 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments

Various studies have examined the effects that the litigation for punitive damages has had on people who were originally affected by the spill. A survey of Cordova residents in 2006 (Picou and Martin) determined that 38% of respondents were involved in litigation against Exxon and 74% of these expected to receive money, and that 23% agreed that the litigation process continued to be a source of stress. The study showed higher levels of stress and depression for residents involved in litigation than for those who were not involved. The study concluded that “the inability of the legal process to produce a timely resolution to damages experienced by survivors of the EVOS has actually resulted in a secondary disaster which may be more significant than EVOS” (Picou and Martin 2006: p 38). The authors further postulate that the eventual payment of claims for punitive damages could actually result in another secondary disaster. A later study by Picou (2009) observed that, although EVOS had substantial adverse effects on residents of the region, the litigation associated with EVOS became a separate stressor that exacerbated the effects of the spill. He concluded that “the twenty years of litigation have exerted a tremendous toll on the communities and residents of Prince William Sound” and “drained their social and cultural capital, and destabilized residents’ sense of community and trust in others” (Picou 2009: p. 20). He postulates that, as a result of the final awards for damages, the community will not recover over the life of the plaintiffs.

5.2.10.5 Summary on Compensation and Litigation Thus, a key factor in the recovery of the human environment following a marine oil spill is the length of time required to settle all claims for damages. In the case of the Exxon Valdez, the complicated process of pursuing claims through the courts took 20 years to resolve and left many plaintiffs believing that the final award was insufficient to offset the damages they had experienced, or to meet their expectations of remuneration through the class action process. In contrast, a review of the Selendang Ayu spill concluded that there would be no substantial long-term effects associated with litigation because the civil and criminal cases were concluded relatively quickly (Impact Assessment Inc. 2011). The settlement of federal criminal charges was resolved nearly three years after the spill occurred, and resulted in a criminal penalty of $10 million. The State of Alaska sued for costs and damages and the resulting agreement, reached 4.5 years after the spill, was for nearly $845,000. Individual fishermen were compensated based on claims for revenues from the lost fishing season and, by submitting claims, fishermen forfeited their options for litigation. According to Impact Assessment Inc. (2011), a “collective decision not to litigate was made in part because of the fishermen’s direct and indirect familiarity with the protracted litigation and settlement processes associated with the Exxon Valdez oil spill.” Additional damages may also be assessed following completion of a Natural Resources Damage Assessment for the Selendang Ayu spill but, as of 2011, this had not yet been completed. Not all the claims related to the Selendang Ayu spill have been resolved, however. The Qawalangin tribe submitted a formalized claim for lost subsistence opportunities to the National Pollution Funds Centre, which can use the Oil Spill Liability Trust Fund (OSLTF) to pay for damages not paid by the party responsible for the spill. While the amount of the claim was

July 2012 Page 5-89

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 5: Marine and Estuarine Environments developed using an approach that represented the Qawalangin tribe’s use of the area in terms of trips, the claim was rejected because it did not describe the actual subsistence use of natural resources that were affected, did not describe the nature of the damages suffered or the efforts taken to mitigate these losses, or provide a reasonable replacement cost (Impact Assessment Inc., 2011). The Qawalangin tribe, which had participated cooperatively in the Unified Command and directly in the food safety program did not choose to appeal the decision or to submit a revised claim. It should be noted that while the American experience regarding the effects of litigation is instructive, caution must be exercised in extrapolating that experience to a Canadian spill event, which would be subject to different legal processes, rights and entitlements. As explained in the Northern Gateway Application (and Information Responses), Canada’s Marine Liability Act regime is intended to avoid disputes regarding liability and to effectuate efficient settlement of claims through mandatory insurance and industry-funded pollution compensation fund programs.

Page 5-90 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

6 Terrestrial and Freshwater Environments The literature addressing recovery from oil spills in terrestrial and freshwater environments is sparser than that for spills in marine waters. Terrestrial industrial areas, such as, storage tank farms, generally have containment basins and walls that prevent release of oil to the natural environment. Clean-up within such containment is an essentially complete removal of the spilled oil. Similarly, for some terrestrial oil spills, the contaminated soil and vegetation is commonly removed and replaced with clean soil and replanted vegetation. The time course of recovery in such circumstances follows the time taken for the replanted vegetation to become established. The first two subsections in this section address cases in which oil leaves containment or is not completely removed and flows over the land. Some of the case studies for soil and vegetation are based on experimental spills of spills where portions of the land were not cleaned to examine recovery without clean-up. The remaining subsections address recovery in cases during which spilled oil enters freshwater bodies such as streams, rivers, ponds, and lakes.

6.1 Biophysical Environment

6.1.1 Soil Quality and Terrestrial Vegetation Surface soil generally consists of a complex matrix of air, water, and both organic and inorganic materials of different sizes and shapes. All of these elements play a role in supporting an equally complex and diverse ecosystem that is made up of microbes (including bacteria, fungi, and moulds), soil invertebrates and plants. This diversity allows soil-borne organisms as a whole to occupy a broad range of niches, and tolerant species are quick to colonize an area that has become less favourable for their competitors. This is generally the process by which the system recovers from contaminant insults, such as oil spills. When oil is spilled on soil, it interacts with the matrix in a number of ways, which either individually or in combination, can result in deleterious effects to the soil biota. In addition to being toxic via direct contact with plants, invertebrates and microbes, the oil also has the potential to interact in other ways by: • occupying the pore space between the soil particles thus excluding air and water • coating soil particles and creating a hydrophobic layer that prevents the absorption and • retention of water • retarding subsequent exchange of air and water with the soil matrix and leaving the soil deficient in both of these components along with nutrients that would otherwise be leached into the soil from the surface • increasing the carbon content of the soil and subsequently promoting microbial growth that depletes oxygen and available nutrients.

July 2012 Page 6-1

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Although the latter point presents a situation that is self-regulating, it still has the potential to maintain the affected soil in a relatively anoxic state. Likewise, limitations on the availability of water, nitrogen, carbon dioxide and other nutrient, leaves the soil with a substantially diminished capacity to support the more sensitive indigenous biota.

6.1.1.1 Case Studies for Soil and Vegetation Few studies have investigated the effects of petroleum hydrocarbons (PHC) spills on soil-borne organisms for extended periods of time, and these have focused on plants and the associated microorganisms. Of the existing studies, most have been performed in arctic and sub-arctic areas of Canada and Alaska, an environment with a short growing season and low annual temperatures (Collins et al.1994; Seburn et al.1996, Freedman and Hutchinson, 1975; Harper and Kershaw, 1996); however, the understanding gained from these studies can be extrapolated other climates. Wein and Bliss (1972) looked at the recovery of vegetation in three test plots in northwestern Canada just to the east of the Mackenzie Delta where sweet crude oil was purposely applied at rates sufficient to saturate the soil profile. The affected plant communities included black-spruce- alder-heath, to medium shrub-alder-heath, and sedge-cotton grass-heath. In each plot, the oil completely destroyed the actively growing plants. After one full growing season, total plant recovery was between 20 and 55%, based on ground cover of new growth. There was substantial regrowth from the latent buds of the dwarf shrub specie, as well as by the sedge species. In contrast, the lichens showed no recovery and only one species of mould showed regrowth. Unfortunately, the study was limited to only the one year, and no information was collected to determine if the recovery continued to completion at the same rate. Certainly one of the best studied experimental oil spills took place in 1976 in an open black spruce forest within the Caribou-Poker Creeks Research Watershed just north of Fairbanks, Alaska (Table 6.1). Each of three study plots was 10 m x 50 m with the long axis downslope. One plot was the control; the second, the site of a winter release; and the third, the site of a summer release. Almost 8,000 of hot Prudhoe Bay crude oil were released into each experimental plot. No clean-up activities or site remediation was carried out. As noted in the next section (6.1.2.2), remediation measures can be used to accelerate site recovery.

Table 6.1 Summary Table for the Caribou-Poker Oil Spill Oil Spill Name Caribou-Poker Location Fairbanks, Alaska Year 1976 Oil Type Crude Specific Name Prudhoe Bay Crude Volume (metric tonnes) 7 Platform Experimental Environments Affected Terrestrial

Page 6-2 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Table 6.1 Summary Table for the Caribou-Poker Oil Spill (cont’d) Valued Ecological Component (VEC) Studied • Microbial Community • Soils • Vegetation Summary: In February and July of 1976 two experimental oil spills (each at 7 metric tonnes of crude oil) were conducted on the Caribou-Poker Creeks Research Watershed, (CPCRW) 48 km northeast of Fairbanks, Alaska USA. The purpose of the spill was to examine the potential effects of an oil leak from the concurrently developed Trans-Alaska Pipeline. The spill covered approximately 320 km of affected land area. There was no attempt to clean up the oil spill, thus making it ideal for long term studies. Studies of vegetation, soils and microbial communities were conducted that documented recovery 25 years after the spill. For both of the spill plots, the vast majority of the vegetation including the spruce trees, shrubs, mosses, and lichens were destroyed and after 25 years, none of the black spruce trees had regrown. Oil continues to be degraded by an active microbial population that has remained acclimated to degrade hydrocarbons, but the combination of low rates of nutrient turn-over, and a short thaw season have resulted in high PHC concentration for decades. No clean-up or degradation enhancements were applied, but it was concluded by Braddock et al. (2003) that natural weathering processes will eventually remove much of the hydrocarbons. Following her work on the plant and microbial communities within the test plots, Garron (2007) concluded that the two experimental plots would never recover, but this is more a statement of the definition of recovery than the state of the ecosystem. Black spruce trees were particularly sensitive with nearly 100% mortality and no regrowth. This was also true of the mosses, lichens and shrub understory made up of resin birch, Labrador Tea, blueberry and willow (Collins et al. 1994). However, cotton grass tussocks that survived the initial spill have recovered vigorously, which is thought to be the consequence of competitive release (Collins et al. 1994). Much of the originally affected areas has been revegetated with the cotton grass and the new plant and microbial communities continue to thrive (Lindstrom et al. 1999; Collins et al. 1994; Sparrow et al. 1978).Hutchinson and Freeman (1978) found that when fresh unweathered Norman Wells crude oil and diesel oil was applied as a low-volume spray, an herbicidal effect was noted with all green tissue that came into direct contact with the oil. However, the initial recovery was rapid with regrowth shoots appearing within weeks of the first growing season, and continuing over the following 3 to 4 years. Some species survived as under-ground rhizomes and also made a re- appearance. Further recovery was progressive and mainly occurred from regrowth of surviving plants but also included vegetative reinvasion from unaffected peripheral areas. Only limited seedling establishment was observed for vascular plants and no black spruce recovery was noted (Hutchinson and Freedman 1978). Racine (1994) agreed that where surface soil is lightly oiled or if the oil is subsurface, then continued survival and recovery of the plants is possible, but in areas of heavily saturated surface soils, long-term recovery is expected to be slow except for species such as the cotton grass. In absence of clean-up or remediation efforts, Hutchinson and Freeman (1978) predicted that a minimum of 20 growing seasons would be needed to re-establish total growth cover. It was also noted that the species composition of the recovered community would differ from that of the pre-spill community due to the different tolerances to the residual hydrocarbon.

July 2012 Page 6-3

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Seburn et al. (1996) investigated the effects of a simulated pipeline spill in a subarctic upland black spruce forest located near Tulita, Northwest Territories. Within the first growing season post-spill, an approximate 70% decline in total plant cover was observed in the most heavily oiled areas. The lightly oiled area showed no change from the control plot. Where the soil was saturated with water (low lying ditch areas), even the higher oil application caused no significant decrease in plant recovery cover. The authors hypothesized that the water kept the soil from being saturated with the oil and thus little was retained to inhibit subsequent plant growth during the post-spill growing seasons. Seburn et al. (1996) concluded that the recovery of the ecosystem was dependent on the plant species and the rate of oil treatment. For example, in all but the most heavily oiled areas, mosses became more abundant by the second growing season. By the third growing season, cover of 17 taxa was affected by the oil, particularly at the higher concentrations of crude oil. Lichens, native grasses, shrubs and forbs appeared to be most negatively affected with little or no recovery. Mosses, agronomic grasses and sedges were found to be more tolerant and in the case of the sedge grass Carex spp. actually thrived within the oiled areas, with considerable revegetation being observed. Another study (Belsky 1980), examined a sub-alpine meadow located near Mt. Baker, Washington over a nine year recovery period following a diesel fuel spill. Plant cover within the area of perturbation decreased to 1% from the 80 to 100% that had been present pre-spill. Only a species of shrub (pink mountain heath), a sedge (Carex lenticularis) and a species of moss (Rhacomitrium sudeticum) survived. Seedlings of sub-alpine species started to re-appear after one year, with other plant species following after two to four years. Nine years after the spill, 5 to 20 % of the vegetation ground cover had returned. When the natural re-vegetation of soils historically affected by PHC were examined at a number of sites in aspen parkland and mixed grassland ecoregions, located in southern , it was shown that some plant species and functional groups are tolerant of the altered soil conditions. Robson and colleagues also found that, in general, contaminated land had significantly lower total vegetation cover and different species composition than uncontaminated plots. This indicates that the contaminated sites are at an early stage of ecological succession (Robson et al. et al. 2004). Compared to control areas, mycorrhizal, woody and vegetatively reproducing species, and species using birds or unassisted means for seed dispersal were significantly less common on contaminated plots (Robson et al. 2004). Self-pollinating species were more common. The most abundant plants within the re-vegetated areas included a species of annual forb and perennial grasses (Robson et al. et al. 2004). Evidence also suggests that the time of the year that the spill occurs will affect the extent of susceptibility among vegetation. The artificial oiling of soils during different seasons and at three separate sites within the Mackenzie delta was used to study the initial and long-term effects on the survival and re-invasion of Low Arctic plant species (Wein and Bliss 1972). A total of five plant community types were represented including black spruce-alder-heath (Inuvik), medium shrub-alder-heath and sedge-cotton grass-heath (Tununuk Point), and willow-birch-heath, and

Page 6-4 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments sedge (Tuktoyaktuk). The oil was either a light gravity sweet crude or a hot crude and was applied in either the spring, summer or winter to simulate the range in weather conditions. The toxic effect of the oil on the plants was observed shortly after release with many of the actively growing species being completely destroyed. However, even at the highest application rate, some of the plants survived depending on the time of treatment. On the spring-treated plots, 14 species of vascular plants and one species of moss showed signs of regrowth by the end of the first summer. However, in no case did the regrowth cover more than 5% of any plot. In contrast, the summer treated plots showed no evidence of plant recovery during the first treatment year and observations during the following spring indicated that the winter application also severely damaged the vegetation (Wein and Bliss 1972). Following one full growing season after oil treatment, the mean recovery of total vascular plant cover ranged from 55% in the sedge community to 33% in the willow-birch-heath to slightly below 20% in the black spruce-alder-heath and medium shrub-alder-heath community types. Of the individual species, Carex spp. (sedge) and the dwarf shrubs showed the greatest recovery. Lichens showed no recovery in any of the plots while mosses showed only a small amount. In general, plant recovery was greatest following the spring treatments. It was hypothesized that the spring ice protected the plant roots from the oil and that by the time the soil melted to the rooting zone, the toxic volatile component of the crude oil had been lost to evaporation. During the summer application, the fresh oil was able to fully penetrate the soil and make full contact with the plant, thus causing the most damage. Interestingly, the winter spills caused more damage than expected (Wein and Bliss 1972). However, the authors predicted that from the results to date, regrowth of dwarf birch, willow and heath shrubs will be considerable within 3 to 5 years, depending on the toxicity of the crude oil (Wein and Bliss 1972). In addition, several studies indicate that if oil exposure is limited to aerial parts of the plant (stems and leaves) the recovery will progress faster as new growth can re-emerge from surviving stems or roots (Holt 1987). Finally, the effects of diesel oil appeared to be more toxic to plants than crude oil (Holt 1987).

6.1.1.2 Enhancing Recovery of Soil and Vegetation Several studies have examined the effectiveness of remediation techniques in accelerating recovery processes. For large, remote areas suitable remediation approaches are limited to those that can be executed in-situ. Bioremediation, the enhancement of the growth of endogenous PHC-degrading soil microbes through the addition of nutrients, has been widely used. Fertilizer addition has been shown to increase bioremediation rates (first-order rate constant of 0.0033 days -1 with fertilizer and 0.0020 days -1 without fertilizer) and accelerate ecosystem recovery in soils affected by crude oil (Sublette et al. 2007; McKendrick and Mitchel 1978). Another biological remediation strategy, phytoremediation, utilizes plants to degrade, contain and sequester soil contaminants. Phytoremediation is performed with plant species that have demonstrated the ability to increase PHC degradation. Like fertilization, plants accelerate PHC removal rates by also encouraging the growth of soil microbes, including PHC-degrading microbes, with organic nutrients released from plant roots. Often the microbial numbers near the

July 2012 Page 6-5

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

plant root are as much as one to two orders of magnitude higher than those of unvegetated soil and it is these microbes that are thought to be predominantly responsible for PHC removal (Siciliano and Germida 1998; Robson 2004; Kamath et al. 2004). Phytoremediation has been particularly successful and is thought to be a more effective way to remediate PHC than the use of bioremediation, where microbes alone degrade the contaminants (Siciliano and Germida, 1998; Gurska et al. 2009; Huang et al. 2005). Studies investigating fertilizer addition in phytoremediation applications indicated that it accelerates PHC degradation; however, addition of fertilizer to soil (bioremediation) does not produce similar results (Merkl et al. 2005). Several trials with PHC contaminated soils demonstrated that certain plant species (mixes of grasses and legumes) significantly remediated PHC affected soil and reduced toxicity (Schwab and Banks 1999; Banks et al. 2003). Often toxicity of PHC hinders phytoremediation applications as considerable plant growth and microbial growth is needed for degradation. Even with the addition of nutrients, high enough levels of microbial and plant biomass may not be achieved to observe significant remediation. One way to address this issue has been to utilize naturally occurring Plant Growth Promoting Rhizobacteria (PGPR) to enhance plant growth and remediation (Gurska et al. 2009). Studies at a contaminated land farm site indicated that as much as 70% of PHC contamination can be degraded over the course of three years with PGPR enhanced phytoremediation (Gurska et al. 2009); PGPR-treated plants remediated 30% more than untreated plants. The use of PGPR could be targeted to specific sites and ecosystems; by isolating PGPR from a particular site and utilizing it in phytoremediation applications to enhance ecological restoration. Although typically phytoremediation involves the planting of non-native species that have been demonstrated to increase PHC degradation in soils, interest in native grasses is rising as simultaneous remediation and ecological restoration would be most desirable (Nedunuri et al. 2010). The natural revegetation or natural attenuation that will occur at PHC contaminated sites will contribute to remediation of the affected site. The use of fertilization to expedite the natural revegetation of anthropogenically disturbed environments might be beneficial when shorter growing season and low nutrient availability are of concern (Deshaies et al. 2009; Nedunuri et al. 2010). Initial studies show that revegetation with indigenous plant species with a single dose of mineral fertilizer would be preferred; although evidence also suggests that organics may be of more benefit than mineral fertilizers (Deshaies et al. 2009).

6.1.1.3 Summary on Recovery of Soils and Vegetation from Spills The recolonization of PHC-affected soils by the indigenous plant species varies from a few weeks after the spill, to timelines on the order of decades. The most hydrocarbon tolerant plant species will be first to reappear in succession (Robson et al. et al. 2004) but the recolonization and recovery will depend on the rate of weathering of PHC in soil that renders the PHC less phytotoxic, and returns both the soil structure and nutrient levels to that which allows for plant germination and growth (Belsky 1980). The natural weathering of PHC in soil consist of physical, chemical and biological processes such as volatilization, photo-oxidation and natural biodegradation by soil microorganisms (Wang and Fingas 1997).

Page 6-6 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Therefore, recovery does occur depending on the degree of soil oiling and environmental conditions, but can take decades unless site clean-up and remediation measures are employed. A variety of remediation approaches are available, and research is demonstrating that combining phytoremediation with the informed use of fertilizers and other methods for supporting microbial activity is effective in increasing the extent and rate of biodegradation of PHC in the soil.

6.1.2 Groundwater Quality Groundwater is ecologically important for sustaining rivers, wetlands and lakes, as well as being an important source of potable water for humans. Seepage of contaminated groundwater may affect the aquatic environment. When oil is spilled into soil, it will migrate vertically downwards under gravitational and capillary forces. As the oil continues moving downwards, globules of the oil become trapped in the soil, unable to overcome the capillary forces. Depending on the size of the spill, the oil may be retained in the soil, or if sufficient oil is spilled, the oil may reach the water table, spreading along the water table and migrating in the direction of groundwater flow. Although oil is considered a non-aqueous phase liquid (NAPL), water-soluble chemicals within the oil will dissolve into water percolating through the soil or, if oil has reached the water table, directly into the groundwater. The result is a plume of contaminated groundwater emanating from the spill zone in the direction of groundwater flow. The extent of effects on groundwater depends not only on the volume of the spill, but also on the characteristics of the oil (density, composition), the hydrogeological characteristics of the spill site (e.g., permeability, depth to groundwater, groundwater velocity), and the length of time elapsed prior to clean-up activities. The natural attenuation processes that support groundwater recovery of spilled oil include: • Biodegradation – the breakdown of chemicals by microbial organisms through metabolic or enzymatic action • Sorption – chemicals can sorb to the natural organic material that is present in the mineral grains of the aquifer (i.e., the soil or bedrock that contains the groundwater) • Groundwater dispersion – as dissolved chemicals are transported with the groundwater, mechanical mixing of the water results in a spreading of the groundwater plume, which increases the plume size while decreasing the concentration of the contaminant • Volatilization – volatile oil components may leave the oil and the dissolved plume through volatilization and biodegradation in, and diffusion through, the unsaturated soil above the water table The plume extent is stabilized and controlled by a balance between the solute transport processes advancing the plume away from the spill (groundwater advection and dispersion) and the biodegradation process. Over time, as concentrations in the source area are reduced, biodegradation processes will ultimately shrink the plume. While natural attenuation processes

July 2012 Page 6-7

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

occur to varying degrees at all petroleum hydrocarbon spill sites, NAPLs are not readily subject to natural attenuation in the short term.

6.1.2.1 Enbridge Line 6b Kalamazoo River, Michigan In July 2010, more than 800,000 gallons of crude oil flowed into Talmadge Creek, which is a tributary of the Kalamazoo River (Table 6.2). The spilled oil was eventually contained at Morrow Lake, which was more than 30 miles downstream from the spill. At the time of the spill, Talmadge Creek and the Kalamazoo River were between 25- and 50-year levels due to the rain that had fallen during the previous days (MDCH 2012). Because the river and creek were at high water levels, oil flowed into overbank areas, wetlands, and floodplains. A hydrogeological evaluation and a drinking water well sampling program were implemented to address concerns that the spill may affect drinking water supplies.

Table 6.2 Summary Table for the Enbridge Line 6b Oil Spill Oil Spill Name Kalamazoo River Location Michigan, USA Year 2010 Oil Type Heavy Crude Specific Name none Volume (metric tonnes) 3,247 Platform Pipeline Environments Affected Freshwater Valued Ecological Component (VEC) Studied • Birds • Fish • Macroinvertebrates • Reptiles • Sediment • Water Quality Summary: On July 26, 2010 a pipeline transporting oil through south-central Michigan ruptured and spilled approximately 3, 200 metric tonnes of crude oil. The stream that was directly affected was Talmadge Creek a tributary to Kalamazoo River which flows into Lake Michigan. Approximately, 56 km of stream habitat was affected including the Kalamazoo River. The US EPA had reported that 1.1 million gallons of oil had been recovered from initial cleanup efforts. Studies of a number of VECs were followed documenting recovery status. Over 30 monitoring wells were installed in areas where oil remained in the floodplain or river banks, and the potential for groundwater effects was considered greatest. It was determined that in most areas, groundwater flows towards the river, and therefore groundwater used for drinking water would not be expected to be affected by the remaining oil (Enbridge 2010). Crude oil constituents were not present in groundwater samples collected from the monitoring wells, indicating that residual oil had not migrated to groundwater, even in areas where water flows from the river to the groundwater (Enbridge, 2010).

Page 6-8 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

The drinking water well sampling program was initiated for wells located within 200 feet of the high water mark from the July 27 (in 2010) flood event. As of August 2011, more than 150 wells had been sampled, and over 600 individual water samples had been analysed (MDCH 2012). Only two oil-related inorganic chemicals, nickel and iron, were found in private drinking water wells; however, similar concentrations of iron and nickel were previously detected from wells in Calhoun and Kalamazoo Counties and are likely naturally occurring metals (MDCH 2012). No oil-related organic chemicals were found in the drinking water samples, and the results do not indicate that petroleum products are present in the groundwater (MDCH 2012). Based on the multiple lines of evidence (i.e., groundwater flow direction, monitoring well chemistry, drinking water well chemistry), crude oil effects to groundwater from the Kalamazoo spill were not evident in the study results.

6.1.2.2 Bemidji, Minnesota The crude-oil spill site near Bemidji is one of the better characterized terrestrial crude oil spill sites in the United States, and has resulted in over 200 publications (USGA 2012). The spill occurred on August 20, 1979 when a crude-oil pipeline burst, spilling about 1,700,000 L (about 10,000 barrels) of crude oil. After clean-up efforts were completed in 1980, about 400,000 L (about 2,500 barrels) of crude oil remained in the unsaturated soil and near the water table (USGS 1998). In 1983, the U.S. Geological Survey Toxic Substances Hydrology sponsored the establishment of a long-term, interdisciplinary research project at the site. From 1983 to 1999, scientists studied the effects of the physical, chemical, and biological processes driving the degradation and transport of crude oil under natural, undisturbed conditions (Essaid et al. 2011). In 1999, a 5-year pump-and-skim remediation effort to remove the NAPL (crude oil) source was initiated in response to a mandate from the Minnesota Pollution Control Agency; however, research at the site continues to focus on methods for measuring and investigating in situ properties and natural processes (Essaid et al. 2011). As of 2011, the oil persists at the site as a separate fluid phase (NAPL), as dissolved petroleum constituents in groundwater, and as vapors in the unsaturated zone (USGS 2012). As of 1996, the leading edge of the oil floating on the water table had moved about 40 m downgradient since the spill. Mass loss rates of crude oil at different locations range from 0 to 1.25 percent per year. The total loss of oil mass was estimated at 11 percent from 1979-89 (USGA 1998). The dissolved plume has not migrated as far as predicted considering the groundwater flow velocities and sorption constants for these compounds (Baedecker et al. 1993). As of 1996, the leading edge of the plume of groundwater had moved only about 200 m downgradient, whereas advective flow of groundwater since the spill has been about 500 m. It has been concluded that primary reason is that hydrocarbons have biodegraded under oxic and anoxic conditions, with 46 percent of the dissolved contaminants degraded after 13 years (USGS 1998). Of this, aerobic degradation accounted for 40 percent of the total dissolved organic compounds degraded and anaerobic processes accounted for 60 percent (USGS 1998).

July 2012 Page 6-9

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Monitoring of the soil gases in the area of the spill has also been on-going. As of 1985, the leading edge of the plume of hydrocarbon vapours in the unsaturated zone was about 150 m downgradient (USGS 1998). By 1997, the plume of hydrocarbon vapours had receded to about 75 m downgradient. This recession has been attributed to aerobic biodegradation (USGS 1998). Currently, as much as 1.2 m of crude oil is floating on the water table (USGS 2012). An estimate of a time for full recovery has been placed at 110 years (Revesz et al. 1995). This time is considered a minimum estimate as the observed rate of degradation may slow down as weathering causes the more volatile and reactive components to leached out (Revesz et al. 1995, Essaid et al. 2011).

6.1.2.3 Summary on Groundwater Quality The reality is that for most terrestrial spills, the amount spilled is small and standard spill response activities are effective at removing the source (NAPL) before groundwater becomes affected. Sometimes even large volume spills may have no observed effect on groundwater (Kalamazoo). In the absence of an on-going source, dissolved plumes tend to be of limited extent and may be addressed effectively through Monitored Natural Attenuation (MNA). If on the other hand, NAPL reaches the water table and cannot be removed, natural attenuation has limited effect on the NAPL, and the NAPL becomes an on-going source of a dissolved plume (e.g., Bemidji). Of the two crude oil spills examined, no effects to groundwater were observed in one and partial recovery as the return of dissolved petroleum hydrocarbon constituent concentrations to background levels have been observed after 30 years.

6.1.3 Freshwater Benthic Organisms Organisms associated with submerged substrates in lake and riverine ecosystems are collectively referred to as the benthos (Horne and Goldman 1994). Examples of these organisms include aquatic insects, crustaceans, molluscs, and worms among others. Organisms that live on the bed surface are referred to as epibenthic organisms, while those that burrow into the substrate and live beneath the surface are referred to as infauna. Benthic organisms typically have little commercial value; however, they are critical components of aquatic ecosystems because of their role in converting non-living organic matter into energy resources accessible to other aquatic organisms. This is especially true in smaller forested stream ecosystems that may derive the majority of their energy budgets from terrestrial sources such as leaf and needle litter (Vannote et al. 1980, Allan 1995). Benthic organisms are functionally adapted to process fine and coarse particulate organic matter and are commonly classified by functional feeding groups (Table 6.3).

Page 6-10 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Table 6.3 The Feeding Roles of Invertebrate Consumers in Running Waters Feeding Roll Food resource Feeding mechanism Examples Shredder Non-woody CPOM, Chewing and mining Several families of primarily leaves; and Trichoptera, Plecoptera associated microbiota, and Crustacea; some especially fungi Diptera, snails Shredder/gouger Woody CPOM and As above Occasional taxa among microbiota, especially Diptera, Coleoptera, fungi; primarily surficial Trichoptera layers are utilized Suspension feeder/ FPOM and microbiota, Collect particles using Net-spinning Filterer-collector especially bacteria and setae, specialized Trichoptera, Simuliidae sloughed periphyton in filtering apparatus or and other Diptera; some water column nets and secretions Ephemeroptera Deposit feeder/ FPOM and microbiota, Collect surface Many Ephemeroptera, Collector-gatherer especially bacteria and deposits, browse on Chironomidae, and organic microlayer amorphous material, Ceratopogonidae burrow in soft sediments Grazer Periphyton, especially Scraping, rasping and Several families of diatoms; and organic browsing adaptions Ephemeroptera and microlayer Piercing Trichoptera; some Macrophytes Diptera, Lepidoptera and Coleoptera Hydroptilid caddis larvae Predator Animal prey Biting and piercing Odonata. Megaloptera, some Plecoptera, Trichoptera, Diptera and Coleoptera Source: from Allan 1995 As described previously, benthic organisms are coarse and fine particulate organic matter into other forms of energy. They are also important consumers of autotrophs such as periphyton and macrophytes. Some of these organisms are predatory and use a number of techniques to capture prey including those that ambush prey, as well as organisms that trap prey in elaborate nets or those that lay mucus traps (Allan 1995). Benthic organisms are important prey items as reflected by the fact that most North American freshwater fish are invertivores (Allan 1995). Benthic macroinvertebrates have a high degree of site fidelity, have complex life cycles, are relatively easy to sample and identify, and are organisms with a variety of trophic positions and pollution tolerances. Consequently, benthic invertebrate communities have long been considered excellent indicators of the overall ecological integrity of aquatic ecosystems (Barbour et al. 1999). The following section describes a number of case studies on the recovery of the benthic community following oil spills.

July 2012 Page 6-11

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

6.1.3.1 Pine River, British Columbia Spills on the Pine River occurred in British Columbia in 1994 and 2000, and are described more fully in Table 6.4. Impairment of benthic invertebrate communities was observed for at least 120 km downstream of the 2000 spill (de Pennart 2011). In eight of the ten sample stations, relative abundances were less than 10 percent of abundance observed at the upstream control site. By July of 2001, benthic invertebrate relative abundances at sample stations between 1.6 and 37.3 km downstream of the spill location were higher than relative abundances observed at the upstream control site (de Pennart 2004). Relative abundance did increase from 2000 levels at sampling stations between 60.4 and 121.7 km but did not reach levels observed at the upstream control sites. In one case, relative abundance was less than half of that observed upstream of the spill. Relative abundance declined substantially at the furthest sampling point from the spill (186.6 km). Overall, the data suggest that benthic invertebrates had recovered in much of the Pine River by July of 2001, less than one year from the date of the spill.

Table 6.4 Summary Table for the Pine River Oil Spill Oil Spill Name Pine River Location British Columbia, Canada Year 2000 Oil Type Light Crude Specific Name none Volume (metric tonnes) 704 Platform Pipeline Environments Affected Terrestrial, Freshwater Valued Ecological Component (VEC) Studied • Aquatic macroinvertebrates • Fish • Sediment Summary: Two spills were recorded for the same reach of the Pine River in British Columbia. The first spill occurred on August 18, 1994 and released of 22 metric tonnes of gasoline and 20 metric tonnes of diesel fuel from a tanker truck into the Pine River. Description of the intervention measures used for the 1994 spill was not readily available to the authors of this report. The most recent spill occurred on August 1, 2000 as a result of a pipeline rupture. Approximately 704 metric tonnes of crude oil was spilled to the Pine River, with between 41 and 97 metric tonnes of the oil being recovered in the two-month clean-up effort. Oil contaminated woody debris jams were removed from the channel and burned. Other intervention measures were not described in detail.

6.1.3.2 Enbridge line 6b Kalamazoo River Qualitative macroinvertebrate community and habitat surveys were conducted in September 2010 and August 2011 (MDEQ 2012) in Talmadge Creek and the Kalamazoo River. Two control sampling sites were established on Talmadge Creek (T1, T2) upstream of the source of the spill and one control was established on the Kalamazoo River (K1). Sample location T3 was located in Talmadge Creek approximately one mile downstream of the spill point. Three sample sites

Page 6-12 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

were located in the spill zone of the Kalamazoo River at approximately 2.75 (K2), 7.25 (K3), and 21.25 (K4) miles downstream of the spill point. Nearly all of the sample sites had been sampled previously at various intervals between 1994 and 2008. Pre-spill taxa richness at the control sites on Talmadge Creek ranged between 16 and 19 (Figure 6.1a). The impact site was not sampled prior to the spill. Taxa richness in the spill zone was considerably lower than taxa richness at the control sites. By August of 2011 the number of taxa increased to 24 and was approximately equal to or greater than taxa richness at the upstream control sites. Community indices scores (MDEQ 1990) fell within the “acceptable” range (-4 to 4) in all three sampling events (Figure 6.1b). However, the 2010 data bordered on a “poor” (-5 to -9) classification while the 2011 data approached “excellent” (+5 to +9) for the same location. The community was numerically dominated (80%) by filter feeders, collector gatherers, and filamentous algae piercers and was characterized as “early successional” species. The authors speculate that removal of trees, shrubs, and herbaceous plants as part of the clean-up potentially increased primary production in the stream and led to the dramatic increase in taxa richness, abundance, and community composition in 2011.

Note: River miles less than 0 denote upstream control sites. Figure 6.1 Benthic Macroinvertebrate Taxa Richness (a) and Community Quality Index (b) by Approximate River Mile in Talmadge Creek By 2011, taxa richness at river mile 2.75 and 7.25 in the Kalamazoo River was greater than or equal to taxa richness at the upstream control site (Figure 6.2a). Taxa richness at river mile 21.25 was lower than the upstream control site but was also slightly lower during the pre-spill sampling. Macroinvertebrate sampling in 2011 indicated “excellent” quality communities were present at river mile 2.75, 7.25, and the upstream control (Figure 6.2b). Community scores were lower at river mile 21.25 but were still “acceptable”. Pre-spill sampling values at this location ranged between 2 and 6 suggesting that variability at this site may be unrelated to the spill.

July 2012 Page 6-13

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Notes: 1) River miles less than 0 denote upstream control sites. 2) Replicate sample sites and/or multiple sample years (pre-spill) are presented as averages. Figure 6.2 Benthic Macroinvertebrate Taxa Richness (a) and Community Quality Index (b) by Approximate River Mile in the Kalamazoo River These data indicate that recovery is occurring in Talmadge Creek and the Kalamazoo River by August of 2011. However, the MDEQ (2012) reports that overall abundance is less than ideal in some of the spill zone sampling sites. This conclusion was reached because it was necessary to count organisms from the entire composite sample rather than a sub-sample as is customary in higher quality streams. In any case it seems clear that some recovery has occurred from 2010 levels. Further sampling in the project area is anticipated in 2012.

6.1.3.3 Cayuga Inlet, New York The spill occurred on November 3, 1997 resulting in the release of 23 metric tonnes of No. 2 fuel oil to the creek. Chemical containment booms were used to collect oil on the water surface but were characterized as only marginally effective. Macroinvertebrate communities were sampled at increasing distances from the spill with 11.8 km being the most distant location. Tributaries in close proximity to the main stem were also sampled and served as reference sites. Sampling occurred within 26 days of the spill and subsequently in the spring and fall of 1998 and in early 1999. Immediately after the spill, invertebrate density at the upstream reference site was approximately 1,000 individuals per 1 minute of sampling. In contrast, the density at 0.7 km below the spill was approximately 100 per sample. Invertebrate densities at the impact site furthest from the spill (11.8 km) were less than 200 per sample and differed significantly from the reference location (~375 per sample). However, the magnitude of difference did not appear to be as great as differences observed at upstream sites. Reduction in the numbers of taxa present at the affected sites was observed in comparison to reference sites. This effect was observed at least 5 km downstream of the spill and lasted for at least three months.

Page 6-14 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

The impact area immediately below the spill (0.7 km) was dominated by the riffle beetle (Optioservus) immediately after the spill. This dominance persisted through the spring of 1998 after which this location was dominated by Ephemerella. In the spring of 1998, autumn of 1998, and spring of 1999, dominant taxa accounted for far more than 50% of individuals observed at the 0.7 km site. The mean proportion of the assemblage consisting of a single taxon was higher at affected sites in all years and at all but one sample station, although the means differed statistically in only 7 of 12 groupings. One year after the spill, there were no significant differences in taxonomic richness between reference-effect pairs. However, in the final year of sampling, there were significant differences in taxonomic richness at the sites nearest and furthest from the spill. Richness at the intermediate distance site did not differ significantly in the reference and affected sites. At the conclusion of the study, among the affected sites, taxonomic richness was highest at the location closest to the spill. By the fall of 1998, invertebrate densities did not differ significantly between effect and reference sites. Further, by 1999 invertebrate density was, on average, higher at two of the three effect stations sampled. Overall density at stations 5.0 and 11.8 remained low in the spring of 1999; however, densities at the reference stations on the tributaries close to these effect sites were also very low in comparison to the upstream reference. A single taxon dominated the fauna immediately below the spill for the entire study. By the final sampling event, significant differences remained in the proportion of dominant taxa at two of the three effect-reference pairs. Although some metrics indicated recovery, the authors concluded that full community recovery may require more than the 15 month duration of the study. Potential reasons for the extended time to recovery include spill volume, propagule supply and seasonal timing. The volume of the spill (26 m3) was large relative to the average discharge in the creek (0.5 – 3.4 m3/sec). The macroinvertebrate fauna were significantly affected at least 11.8 km from the spill origin. There were two main sources of propagules 1) upstream to downstream invertebrate drift and 2) aerial recolonization of adults. This study included two periods of relative dormancy during the winter months but only one period of active colonization.

6.1.3.4 Asher Creek, Missouri In August 1979, an oil pipeline burst spilling 1,000 metric tonnes of crude oil into Asher Creek in southwest Missouri. Six surface skimming siphon dams were deployed along approximately four kilometres of the stream and oil was contained to a large degree in this area. Oil recovery efforts lasted approximately eight weeks and pumps and absorbent pads were used to recover oil collected behind the dams. The volume of oil recovered was not reported. However, Crunkilton and Duchrow (1990) reported that the skimming siphon dams were very effective. Study sites were located 1.4, 3.4, and 7.4 km downstream of the spill area. A reference sampling site was located 0.6 km upstream of the spill location and was intended to represent unaffected conditions. The benthic macroinvertebrate community was sampled 25, 38, 64, 96, 176, 213, 266, 336, 395, 453, and 532 days after spill. Field sampling occurred between September 1979

July 2012 Page 6-15

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments and February 1981. Water quality measurements for dissolved oxygen concentrations, pH, and temperature in the oiled reaches did not differ from measurements at the reference site. Oil sheens were observed at sites 1.4 and 3.4 km downstream of spill location and were present for 336 days at 3.4 km and 453 days at 1.4 km. Stream substrates in these areas were inundated by oil that penetrated the interstitial spaces. Taxa richness and density at the heavy impact site (Site 2) were very low on the date of initial sampling but increased dramatically after only 13 days. However, the taxa were primarily pollution tolerant organisms associated with the Chironomidae, Simuliidae, and Oligochaeta families. Mayflies were absent at this site for 176 days and stoneflies were absent for 64 days. At 266 days after the spill, pollution intolerant organisms were observed in greater numbers and may have been beneficiaries of scouring flows earlier in the spring. Oil did not appear to penetrate substrates at the downstream Site 4. Nor was oiling observed as sheen on the water surface or on the substrates. Metrics of biological quality at Site 4 tracked closely with the reference Site 1 suggesting little or no impairment to the benthic communities. Benthic shredders and pollution tolerant organisms (e.g., Chironomidae) remained unaffected for the duration of the study. Crunkilton and Duchrow (1990) estimated that complete recovery of the benthic biota occurred at approximately 266 days. The data suggest that recovery might have occurred sooner if not for the drought that occurred in the first summer after the spill. The additional stress of warm summer temperatures and low flows apparently interrupted a recovery that was in progress in the early days of the spill. At the time of the spill, the total volume of the spill (1.5 million litres) represented 6% of the total volume present in the stream. Had the initial volume of oil been lower or the discharge higher, reported effects may have been lower. The scouring spring that occurred after the first winter were thought to have contributed to positive responses in the metrics studied and was described as the single most important factor in the dissipation of oil. Asher Creek originates from Cave Spring only 0.6 km upstream from the spill site. Consequently, the abundance of upstream organisms available to colonize affected downstream habitats may have been lower than in surface water driven streams with longer lengths.

6.1.3.5 East Walker River, California In the East Walker spill (Table 6.5), the cumulative number of Ephemeroptera, Plecoptera, and Trichoptera (EPT) taxa in samples collected in 1999 prior to the spill was much higher than numbers observed after the spill on December 30, 2000 (Figure 6.3a). This was true for sites within the impact zone, as well as reference site upstream of the influence of the oil spill. In 1999 the cumulative number of EPT taxa ranged from 15 – 27 in the reference zone and 21 – 23 in the spill zone. Cumulative EPT taxa never returned to pre-spill levels for the duration of the study. Unoiled areas ranged from 2 – 11 in comparison to sites within the spill zone that ranged from 8 – 13. However, in as little as one month after the spill, sites in the impact zone had higher numbers of taxa than the reference locations. There was no discernible change in the cumulative number of EPT taxa over 2001 sampling period (Figure 6.3a). This study also evaluated the percentage of “sensitive” or pollution intolerant EPT taxa ranges for pre-spill reference (3 –

Page 6-16 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

15%) and impact (6 – 15%) sites were similar (Figure 6.3b). The post-spill reference data indicated that the proportion of sensitive EPT taxa present of lower than observed in the pre-spill sampling, perhaps a reflection of the altered flow regime. The percentage of sensitive EPT taxa closely approximated those from pre-spill data downstream of the spill in most cases. In March 2001, the numbers of sensitive taxa within the impact area were much higher between river mile 8.0 and 12.0 than observed in in the other sample events in 1999 and 2001. Neither the cumulative or sensitive taxa metrics exhibited substantial changes with distance from the spill.

Table 6.5 Summary Table for East Walker River Oil Spill Oil Spill Name East Walker Location California/Nevada Year 2000 Oil Type Fuel Oil #6 Specific Name Desulfurized Gas Oil (DGO) / PS 1500 Top Crude Volume (metric tonnes) 11 Platform Truck Environments Affected Freshwater Valued Ecological Component (VEC) • Birds Studied • Fish • Macroinvertebrates • Mammals • Sediment • Water quality Summary: On December 30, 2000 a Tanker Truck overturned and spilled 11 metric tonnes of fuel oil # 6 on a sharp curve at California SR 182 (north of Bridgeport, CA), which entered into the East Walker River. Approximately, 16 km of stream habitat was affected. The estimated volume of recovery from the initial cleanup efforts was approximately 5 metric tonnes of fuel oil. An extensive labor force of 75 people spent 3 months working on cleanup efforts, enduring harsh winter storms with the final cleanup stage ending on March 29, 2001.

The spill occurred in December of 2000 at a time when temperatures were very low resulting in the formation of semi-solid tar-like globs of oil on the bed of the river. In fact, near record low temperatures occurred shortly after the spill and the clean-up was temporarily halted. Agency personnel observed the formation of anchor ice in the channel for the first time in roughly a decade (Hampton et al. 2002). The tar-like formations were very resistant to weathering at low temperatures. Consequently degradation of the oil was delayed until the stream began to warm in the spring months.

Under standard operating conditions, releases from an upstream reservoir were maintained at 20 cfs or higher during the warmer winter months and 30 cfs when temperatures were below 0 C to prevent the formation of anchor ice, freezing of the river, and subsequent winter fish kills. As part of the oil spill response, flow from the upstream reservoir was reduced to 10 – 15 cfs to enable the recovery of submerged oil deposits. Exceptionally cold weather forced a temporary suspension of cleanup activities and anchor ice was documented in the river for the first time in nearly 10 years (Hampton et al. 2002).

July 2012 Page 6-17

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Source: Data adapted from Hampton et al. (2002). Figure 6.3 Illustration of Ephemeroptera, Plecoptera, and Trichoptera (EPT) Taxa by River Mile. Figure a) Represents Cumulative Number of EPT Taxa and b) Represents the Percentage of Sensitive EPT Taxa

Page 6-18 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Average macroinvertebrate abundance upstream of the spill area in January of 2001 was approximately 2,800 organisms per sample site (Figure 6.4). In comparison, average abundance in the impact sites was 254 organisms per sample site. Average abundance at the impact site in March was also far less than at the upstream reference sites. In October, average abundance for the impact sites was similar to values observed at the upstream reference sites. While abundance at the impact site increased slightly, abundance at the reference site was much lower than the prior sampling. It is not currently know whether this change in abundance is related to natural variation or from changes to winter releases from the reservoir.

Legend: US = upstream, DS = downstream, Jan = January 2001, Mar = March 2001, Oct = October 2001. Source: Data adapted from Hampton et al. (2002). Figure 6.4 Mean Macroinvertebrate Abundance by Location and Time Period These lines of evidence jointly suggest that some level of impairment remained at the conclusion of the sampling, but that impairment was observable in both the reference and impact sites. While effects from the spill cannot be ruled out for the impact sites, the impairment at all sites could also be related to the combined system stress resulting from the extreme winter and altered flow regime.

6.1.3.6 Tennessee Pond Trials Experimental trials in Tennessee ponds were conducted from December 1980 through June of 1981 (Cushman and Goyert 1984). Three different levels of synthetic crude oil treatments (15, 75 and 375 ml oil/m3) were applied to 8 plastic lined experimental ponds in Tennessee in July, 1981 for a total of two replicates for each oil treatment (including the control). The ponds measured between 21 and 30 m2 in surface area and were filled with sediment and introduced with macroinvertebrates from pumped water one year prior to the applied oil treatments.

July 2012 Page 6-19

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

The macroinvertebrate community was sampled bi-monthly prior to the oil treatments beginning in December 1980 and extended four months after the treatments (July 1981) into October 1981. The measured parameters of macroinvertebrate community health were immediately negatively affected from the three oil treatments in the first month of post-treatment sampling but showed quantitative signs of recovery three months after treatment. In the first round of sampling in August (after the oil treatment in July), there were significant decreases in species diversity, number of taxa, and abundance of macroinvertebrates in the oiled treated ponds. The greater effect from the highest dosage of oil treatment was reflected in all indices of macroinvertebrate community-level stress (number of taxa, total abundance and abundance by taxon, total biomass and biomass by taxon). There was an especially significant drop in the number of taxa from the medium oil treated pond compared to the number of taxa in the high oil treated ponds. Taxa diversity responded in a similar fashion as the dosage of the oil treatment increased. The total macroinvertebrate abundance was highest in the control ponds and lowest in the high oil dose treated ponds. Pond macrophyte coverage in the high dose ponds was completely eliminated and decreased significantly in one of the medium dose ponds. Recovery was apparent in the species diversity, number of taxa, and total abundance indices recorded from the oil treated ponds. A stimulatory recovery effect of macroinvertebrates was exemplified in the abundance and biomass by taxon in both the medium and high treatment level ponds in the October sampling round. One month after the oil treatment, the mean number of taxa initially ranged from 6.25 in the control ponds to 3.88 in the low oil dose pond, 3.50 in the medium-dose pond and 0.88 in the high-dose pond. However, in three months, the mean number of taxa showed signs of recovery with 4.62 in the low-dose pond, 7.88 in the medium-dose pond and 4.75 in the high-dose pond compared to 6.38 in the control pond. In the three month post- spill sample, insect abundance in one of the high dose replicate ponds was significantly higher than the low-dose, medium-dose, and control ponds.

6.1.3.7 Wolf Lodge Creek, Idaho On June 4, 1983, approximately 72 metric tonnes of unleaded gasoline were spilled into Wolf Lodge Creek, Idaho (Pontasch and Brusven 1988). Five study sites were positioned on the creek to measure the effects of the spill on the macroinvertebrate community in Wolf Lodge Creek, with site 1 (reference) located 420 m upstream from the spill and sites 2, 3, 4 and 5 located downstream of the spill (i.e., the impact sites). Benthic macroinvertebrates were first sampled 17 days after the spill and then sampled at weekly and monthly intervals. Several macroinvertebrate community abundance and diversity indices were used for the statistical comparisons. Although there were immediate negative effects from the spill on the macroinvertebrate community in Wolf Lodge Creek, there was full recovery of the diversity and abundance indices after 16 months.

Page 6-20 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Sampling of macroinvertebrate communities at the reference site and the impact sites 17-days after the spill showed significantly lower macroinvertebrate densities at the impact sites, ranging from only 2-25% of the densities recorded at the reference site. In comparing the macroinvertebrate diversity indices 17 days after the spill, there was a decrease in diversity at sites 4 and 5 downstream of the spill. There were only 36 individuals in 10 taxa and 102 individuals in 11 taxa at sites 4 and 5, compared to 1,843 individuals from 49 taxa recorded at the reference site. Recovery was based on comparing the macroinvertebrate community at each of the affected impact sites (Sites 2-5) with the reference site 1. Macroinvertebrates that have higher drift densities, like mayflies and chironomids, rapidly recolonized the impact sites. Macroinvertebrate community indices of taxon abundance at the impact sites showed recovery to reference levels less than 1 month after the spill. There was no significant difference between macroinvertebrate densities at the impact sites and the test sites 7 months after the spill. Richness recovery took another nine months. Complete recovery of both density and species richness had occurred by 16 months after the spill. The attempted effort to clean spill area on July 9th, 1983, (1 month after the spill) released sediment-trapped hydrocarbons which would have flowed downstream towards the test sites 2, 3, 4 and 5. The reference site (1) was upstream of the bulldozer cleaning area. On July 9th, 1983, (1 month after the spill) a bulldozer entered the stream and attempted to clean the spill area by releasing sediment-trapped hydrocarbons. The degree to which these activities affected macroinvertebrate communities in the work area and in the impact sites in the downstream reaches is uncertain.

6.1.3.8 Summary for Freshwater Invertebrates Recovery of freshwater invertebrates has occurred within 2 years in several cases. Several factors appear to influence the recovery freshwater invertebrates: • the size of the spill relative to the size of the receiving body of water; • the timing of the spill relative to life history; • the amount of oil recovered in clean-up efforts; and • proximity to propagules capable of recolonizing affected areas. External stressors such as extreme drought and/or prolonged cold periods may delay recovery or at the very least complicate interpretation of the data.

6.1.4 Fish and Fish Habitat Freshwater fish play an important role in both human and ecological systems. Fish have traditionally been, and remain, an important food resource for native peoples. Also, recreational fishing is a multi-million dollar industry. Some fish, such as the Pacific salmon, use freshwater for only a portion of their life history and undertake extensive oceanic migrations (salmon are discussed in anadromous fish subsection of the Martine Environment section). Others, while

July 2012 Page 6-21

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

highly migratory, may remain in freshwater. For example, walleye have been observed in locations hundreds of miles from their natal streams (Wang et al. 2007). Still others may reside for an entire lifetime within a few hundred feet of their hatching location. Similar to benthic organisms, fish are functionally adapted to utilize a variety of food resources. Most fish in North America can be classified as invertivores for at least some portion of their life history (Gerking 1994, Goldstein and Simon 1999). Other groups include piscivores (fish that eat fish), planktivores (fish that feed on small drifting organisms), herbivores (fish that feed on aquatic plants), omnivores (fish that feed opportunistically on a wide range of foods), and parasites (fish the feed on the fluids of other fish). Fish may be classified by one of three reproductive strategies (Welcomme et al. 2006): 1) fish that guard their young, 2) those that offer no parental protection, and 3) live bearers. An array of reproductive strategies exists between and among these three general categories. For example, for fish that offer no parental care, some are pelagic spawners that allow fertilized eggs and/or larvae to drift with the prevailing currents. Others broadcast eggs over specific substrates such as coarse gravels, sand, or plants. Fish that offer parental care may also have exhibit specific preferences for certain substrates and many are known to build nests for the care of the young. These strategies have consequences for the vulnerability of life stages to oil exposure. Fish also display differential sensitivities to anthropogenic disturbances. Some species (e.g., fathead minnow) are highly tolerant of polluted conditions while others (e.g., bull trout) are highly intolerant. Feeding guilds, reproductive strategies, tolerance to pollution, and other community indices (e.g., taxa richness, relative abundance, etc.) have been used as indicators of ecological integrity throughout North American (Angermeier and Karr 1986, Barbour et al. 1999, Simon 1999). In the case studies, the recovery of fish populations and recovery of the water column sediment components of fish habitat are examined. Aquatic invertebrates that provide prey to fish were discussed in the previous section.

6.1.4.1 Pine River, British Columbia The Pine River, British Columbia spill is described more fully in Table 6.4. Fish populations in a 50 km section of Pine River were surveyed by snorkeling between 1993 and 2007. Data were collected in 1993 prior to the 1994 spill and approximately three weeks afterward. Data were collected approximately seven weeks after the August, 2000 spill and the Pine River was surveyed three times in 2005, twice in 2006, and twice in 2007. The methods, locations, and number of divers varied between surveys. The 1994 spill was responsible for the confirmed deaths of approximately 1,150 fish. Only 73 fish were observed in the 1994 survey as compared to 2,153 in 1993 and 3,709 in 2005 for the same sections of the river (Goldberg 2011). Density was far lower in the post-spill survey in 1994 (8 fish/km) than after the post-spill survey in 2000 (207 fish/km). Following the August 2000 spill, approximately 1,600 dead fish were documented in a 30 km section of the Pine River. Observed mortality was greatest in a section of the river that was approximately 14 km downstream of the spill (Bustard and Miles 2011). Estimates for total

Page 6-22 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments mortality range between 15,000 and 250,000. Mountain whitefish (Prosopium williamsoni) accounted for 1,124 of the 1,641 mortalities observed. Other fish observed included arctic grayling (Thymallus arcticus), bull trout (Salvelinus confluentus), rainbow trout (Oncorhynchus mykiss), burbot (Lota lota), various suckers, and sculpins. The first post-spill surveys were conducted between September 24 - 28, 2000, roughly five weeks after the spill. Goldberg (2011) concluded that relative abundance in the 2000 survey was lower in comparison to surveys that occurred in 1993 and 2005 in nearby areas. However, the relative proportion of sport fish present was roughly the same as in surveys conducted in other years. One notable exception to this conclusion was that suckers were entirely absent during the 2000 surveys. Further, suckers were only observed in the lower 6 sections of the river surveyed. Goldberg (2011) suggested that recovery of fish communities appeared to be underway less than two months from the date of the August, 2000 spill. The data presented in this analysis point to similarities in relative abundance and species composition in pre and post spill studies. However, it should be noted that abundances presented in Goldberg (2011) combined all of the reaches surveyed by a given consultant and/or year regardless of longitudinal distance from the spill. Therefore, no consideration was given to how fish communities responded along longitudinal patterns for the various sampling intervals. Interpretation of recovery from the August, 2000 spill is complicated in this instance due to differences in field methodologies employed and due to the absence of any kind of experimental control to track inherent system variability. Interpretation of recovery from the 2000 spill is also complicated by the 1994 gasoline and diesel fuel spill and the absence of monitoring between 1994 and 2000. The 1994 data clearly demonstrate impairment from the spill. The state of the fish community and relative degree of recovery in the period between sampling that occurred in 1994 and the spill in 2000 is unknown. Bustard and Miles (2011) suggest that the clean-up effort itself may have caused some damage to the Pine River. For example, oil contaminated woody debris jams were removed from the channel and burned. Summers (2004 as cited in Bustard and Miles (2011)) suggested that as many as 40 structures were removed from the river. Removal of these structures led to channel instability, meander cut-offs, and channel widening. The extent, to which these activities influenced benthic invertebrate and fish community composition and abundance, if at all, is presently unclear.

6.1.4.2 Enbridge Line 6b, Kalamazoo River, Michigan Summary analysis of the impairment and recovery of fish communities from the spill on the Kalamazoo River was not readily available at the time this report was prepared. Continued sampling of fish communities is anticipated in the summer of 2012 (R. Doherty, personal communication). It is possible to conclude that some degree of recovery is occurring because the ban on the consumption of fish from the river was to be lifted on June 21, 2012 (Battle Creek Enquirer.com. 2012).

July 2012 Page 6-23

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

6.1.4.3 East Walker River, California Monitoring of the December, 2000 East Walker River oil spill (Table 6.5) occurred in 2001 and included fish and fish habitat (Higgins et al. 2001, Hampton et al. 2001). Fish habitat was characterized through sampling of the water column and river sediment. Sampling occurred in January, March, and May of 2001. Higgins et al. (2001) found aqueous TPAH concentrations (4,900 ppt) in January 2001 (Figure 6.5a) were higher than the 1,000 ppt threshold associated with mortality of incubating salmonid eggs (Bue et al. 1996, Heintz et al. 1999). However, by May 2001, concentrations returned to levels that posed no immediate threat. Sediment concentrations exceeded designated Threshold Effect Concentration guidelines for sediment quality in freshwater ecosystems as presented in MacDonald et al. (2000). However, by May, concentrations were lower than thresholds for concern. Temperatures were very low at the time that the spill occurred on December 30, 2000. Consequently the oil weathered very slowly and TPAH concentrations in the sediment were low during the January sampling event (Figure 6.5b). As water temperatures increased, weathering of oil also increased causing a dramatic increase in sediment TPAH concentrations. Average concentration for all sites sampled in January was 152.5 ppb and in March exceeded 2,900 ppb. By May of 2001, all but one of the sampled sites was below 350 ppb.

Source: Adapted from data presented in Higgins (2002). Figure 6.5 Total PAH (TPAH) Concentrations in the Water Column (a) and in River Sediments (b) by Distance from the Spill and Sample Date

Page 6-24 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Mean fish densities in 2001 after the spill were far lower than mean densities from pre-spill sampling. In general, densities in Nevada (the furthest downstream reaches of the study area) reported the lowest densities. Data presented in Hampton et al. (2002) for fish densities in the East Walker River exhibit considerable variability over time. However, densities were uniformly low for several species and life stages (Table 6.6). In the case of adult mountain whitefish in Nevada, it is improbable that the observed low densities in 2001 were a function of natural system variability.

Table 6.6 Average Fish Density Pre- and Post-the December 30, 2000 Spill Pre-2001 Data* 2001 Data* Standard % of Pre-2001 St. Devs. from Average Deviation Result Average pre-2001 Data Rainbow Trout in CA 495 410 372 75% -0.3 Rainbow Trout in NV 409 223 29 7% -1.7 Brown Trout in CA 5825 2982 3098 53% -0.9 Brown Trout in NV 587 566 137 23% -1.7 Mountain Whitefish in 546 245 211 39% -1.4 CA Mountain Whitefish in 220 58 71 32% -2.6 NV Young Rainbow Trout 206 160 3 2% -1.3 in NV Source: Data from Hampton et al. 2002 NOTE: *number of fish / mile Whole body TPAH concentrations in samples collected in March of 2001 (Higgins 2001) revealed that suckers had higher concentrations than trout. This likely occurred because of the bottom feeding behavior of suckers in contrast to drift feeding strategies employed by trout. Suckers feed directly off of detritus, algae and other organisms in the benthos that are in direct contact with TPAH contaminated sediments. Trout on the other had feed primarily on invertebrates drifting in the water column. These invertebrates primarily originate from mineral substrates in high velocity, turbulent habitats. Such habitats are more likely to weather and transport oil than the depositional habitats where suckers forage. Despite clear evidence of TPAH uptake in fish tissues, no conclusions were drawn regarding the potential for biological effects from such concentrations. Based on water and sediment chemistry data alone it is possible to conclude that substantial oil degradation had occurred and that recovery was evident in most areas by May of 2001. The fisheries data clearly demonstrate that some impairment occurred,; however, the period of sampling was insufficient to adequately evaluate recovery for fish. The East Walker River Trustee Council (2009) concluded that concentrations no longer posed a threat to fish or to incubating fish eggs.

July 2012 Page 6-25

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

6.1.4.4 Reedy River, South Carolina A petroleum pipeline ruptured and released 3,057 metric tonnes of No. 2 fuel oil (diesel fuel) into the Reedy River. The spill occurred on June 26, 1996. Approximately 94% of the oil was recovered within 12 days. A dam present at the bottom of the study area (37 km downstream of the spill) was a total barrier to upstream fish migration. Another dam approximately 7 km downstream of the spill site also blocked upstream fish migration. Study of the fish community response to the spill began recovery in August 1994 about 4 months after the spill (Kubach et al. 2010). Subsequent sampling was conducted in October 1996, 1997, 1998, 2000, and September-October 2005. Study sites were located 1.8, 14.2, 20.6 and 29.5 km downstream of the spill location and a control sampling site was located just upstream of the spill. The authors reported that fish were nearly exterminated from a 37 km section of the river below the spill. Fish abundance was initially low at the 4 month sampling period but increased substantially at 16 months. Abundance at the 50 month interval was similar among effect and control locations. Species richness was also low initially at all of the affected sites. After only six months, species richness at the affected sites was greater than or approximately equal to richness at the control locations and remained stable for the remainder of the study period. In the early months following the spill, Lepomis macrochirus, a pollution tolerant fish species, accounted for 77% of the total abundance at the affected sites, in comparison to only 12% at the control site. Numerical dominance by this species continued at the affected sites until at least October 1996 (four months after the spill). Initially, metrics of community similarity indicated that all of the affected sites differed significantly from the control location. After only 16 months, Site A (1.8 km downstream of the spill) was very similar in community composition to the control sites. The other affected sites remained clustered together and fell far outside the 95% confidence interval for the control samples. Over time community similarity indices at the affected sites shifted toward the control condition and were nearly convergent in 1997, 2000, and 2005. Pairwise comparisons of control and affected sites fluctuated between significantly different (1996, 1998, 2005) and not significantly different (1997 and 2000). However, the assemblage structure at the downstream locations (B-D) changed little in the period between 2000 and 2005 leading the authors to conclude that recovery had occurred despite the fact that complete congruence in ordination position did not occur. Some indices such as species richness and abundance exhibited signs of recovery in as little as four months. Metrics of community similarity took longer to respond, especially at the sites furthest from the spill. Site A was the impact site closest to the spill area and also to unaffected upstream areas. Consequently, it shifted toward a fish community composition similar to the control sites earlier than the downstream impact areas. One plausible explanation for the differing recovery rate can be attributed to propagules from nearby unaffected areas. The downstream areas were isolated from potential colonists due to the presence of migratory barriers (i.e., dams). The authors concluded that “it appeared that the lower disturbed section (B- D) experienced a slower rate of recovery than the upper section (A), which was near the larger

Page 6-26 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

main stem river assemblage as well as tributaries.” The authors conclude that recovery was complete by the 2000 sampling period, approximately four years after the spill.

6.1.4.5 St. Lawrence River, Quebec Hodson et al. (2002) conducted a study of experimental treatments of oil and nutrients in a freshwater wetland along the shoreline of the St. Lawrence River in St. Croix de Lotbiniere, Quebec. In 1999, 20 5m x 4m plots were treated with one of three different treatments: oil (0.6 L/m2); oil (0.6 L/m2) and nutrients (nitrogen and calcium); and nutrients and no oil. There were also control plots with no oil or nutrients. The objective was to test whether biodegradation of oil could be accelerated through stimulated plant growth. Experimental oiled treated plots were raked to simulate natural activity of wave action on oil penetration into the sediment. The bioavailability of hydrocarbons in the aquatic environment from the oil treatments was measured by purposely exposing juvenile rainbow trout to sediments in oiled plots for four days. The level of assayed enzyme (EROD) activity in the fish acted as a proxy for the bioavailability and exposure of fish to hydrocarbons. The EROD activity of fish in both treated (oil) and control plots (unoiled) was highly correlated to the hydrocarbon concentrations in the treated plots. Therefore, the rainbow trout in oil-treated plots were physiologically affected by the presence of sprayed oil in the plots and sediments. One month after the application of oil, the summed PAH concentrations in the sediments sampled from the oiled plots ranged from 161 - 206 (mean=178.5) micrograms/grams dry weight. There were measurable residual hydrocarbon concentrations in the sediments 15 months after the commencement of the study in both the oil and oil x nutrient treatment plots. There were no significant differences in the hydrocarbon concentrations in the nutrient treated plots. After 15 months, the summed PAH (polynuclear aromatic hydrocarbons) concentrations in the sediments in the oiled plots ranged from 7.4 to 30 (mean = 19.85) micrograms/grams dry weight compared to the unoiled plots and the reference site which measured 1.0 micrograms/grams dry weight. These reductions were most likely due to weathering, evaporation and sediment movement due to currents and tidal river flow. The bioavailability of hydrocarbons in the oil treated plots, as measured by EROD in fish, decreased over the course of the study but was not affected by the nutrient inputs. The bioavailability of hydrocarbons in the oil treated plots decreased by 50% in the first 4 to 5 summer months, 20% over the 6 winter months, and 50% over the second summer months for a total reduction of 80% over the course of the study. EROD induction in fish in oiled and unoiled plots converged over time but the 95% confidence intervals still did not overlap at the conclusion of the study after 455 days. While it is clear that fish continued to exhibit a physiological response to the presence of PAHs, it is not clear if this represented impairment and therefore can only be viewed as partial recovery at this time.

July 2012 Page 6-27

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

6.1.4.6 Summary for Fish and Fish Habitats Recovery of fish communities from oil spills has been observed in as little as four years (Kubach et al. 2010). Signs of partial recovery may be evident after only a few months. Factors that influence the rate of recovery processes include: • the presence of barriers limiting fish movement and recolonization; • the environmental conditions, especially temperature and river flow, during and after the spill; • the life history characteristics of the fish under consideration. For example, pollution tolerant bluegill experienced population growth while bottom feeding suckers recovered less quickly; and • certain types of clean-up methods (e.g., wood removal, flow control) may constitute an additional system stressors and can delay recovery.

6.1.5 Reptiles Bell (2005) studied the effect of crude oil exposure on female snapping turtle and painted turtle fertility, reproductive output, and development of offspring. The study demonstrated that oil exposure did not affect the female adult stage through fertility or relative clutch mass, but did affect the embryonic stages through increased embryonic death and increased prevalence of deformities. Although Bell (2005) did demonstrate increased embryonic death and increased prevalence of deformities, it should be noted that the John Heinz National Wildlife Refuge is subject to pollution from multiple sources other than the spill (Bell et al. 2006). This background pollution places a developmental burden on the life history of these turtles that was exacerbated by exposure to crude oil (Bell 2005). Saba and Spotila (2003) studied the survival and behavior of freshwater turtles after rehabilitation following the SUNOCO spill and found no differences in survival, home range, and temperature preference of 16 oil exposed/rehabilitated (OER) turtles, 18 possibly exposed (PE) turtles, and 32 non-exposed (NE) turtles as measured with temperature sensitive radio transmitters. They concluded that freshwater turtles which had been exposed to oil could be successfully rehabilitated and returned to their former habitat.

6.1.5.1 Enbridge Line 6b, Kalamazoo River, Michigan On July 26, 2010, Enbridge’s Line 6b ruptured near Marshall, Michigan releasing over 3,100,000 liters of crude oil into Talmadge Creek and eventually into the Kalamazoo River (Table 6.2). Boom was erected at several points along the river downstream of the spill site to slow the progress of the oil. Nevertheless, over the course of the next several days after the spill, the oil affected approximately 38 miles of the Kalamazoo River before being stopped at Morrow Lake on the eastern edge of the city of Kalamazoo. Because the Kalamazoo River was at flood stage at the time of the spill, the oil was not confined to the channel and spread out into the floodplain. A

Page 6-28 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

combination of boom, vacuum trucks and tanks, and dredging were used to collect the discharged oil. Rescue and recovery of oiled turtles began within just a few days of the spill and continued into November 2010. A total of 2,670 turtles, comprising eight species, were recovered in 2010 (Table 6.7). The majority (97%) of the turtles recovered were cleaned, treated medically if needed, and released (i.e., rehabilitated). Only a small fraction (2%) were found dead (n=10) or died in care (n=44). A total of 472 turtles were not ready for release by the fall of 2010 due to medical reasons and were therefore overwintered in the project’s Wildlife Care Center during the winter of 2010-2011. Of these, 32 died while in care, 21 became permanent captives due to their injuries, and 419 (89%) recovered and were released in spring 2011, along with 66 hatchlings that were born over the winter.

Table 6.7 Turtles Recovered Following the Enbridge Line 6B Kalamazoo River Spill Total Total 2010 2011 Captures Released Common Name Scientific Name Captures Captures (2010-2011) (2010-2011) Common Map Turtle Graptemys 2030 1659 3689 3592 geographica Common Snapping Chelydra serpentina 188 618 806 793 Turtle Spiny Softshell Apalone spinifera 190 355 545 542 Northern Painted Chrysemys picta 185 245 430 421 Turtle Common Musk Sternotherus odoratus 63 30 93 84 Turtle Blanding’s Turtle Emydoidea blandingii 10 28 38 36 Eastern Box Turtle Terrapene carolina 3 2 5 5 Spotted Turtle Clemmys guttata 1 0 1 1 Source: Enbridge, unpublished data Capture and recovery of oiled turtles resumed in spring 2011 and continued throughout the 2011 active season in an effort to recover turtles missed in 2010 or turtles that had become re-oiled. A total of 2,937 turtles, comprising seven species, were recovered in 2011 (Table 6.5). The majority (98%) of oiled turtles were again cleaned, treated medically if necessary, and released. Only a small fraction (2%) were found dead (n=11) or died in care (n=35). A total of 43 turtles were not ready for release by the fall of 2011 due to medical reasons and were therefore overwintered in the Wildlife Care Center during the winter of 2011-2012. One died while in care, 12 became permanent captives due to their injuries, and the remaining 30 (71%) recovered and were released in spring 2012, along with four hatchlings that were born over the winter.

July 2012 Page 6-29

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Follow-up studies designed to determine the range of effects of the spill and the recovery of the river are on-going and no data are currently available on the success of the recovery of the freshwater turtle fauna. However, the large dataset of recaptures during the 2011 active season does provide some clues. Of 595 turtles caught in 2010 that were marked and released within the affected section of the Kalamazoo River, 252 (42%) were recaptured in 2011, anywhere from six months to over a year after their initial capture. All recaptured turtles in 2011 had successfully hibernated during the winter of 2010 – 2011 and most were in good health when recaptured. These data suggest that turtles that are rehabilitated after an oil spill can be successfully reintroduced back into the spill area and survive, supporting the results of Saba and Spotila (2003). Further investigation is needed into the long-term survival of individuals and populations following an oil spill.

6.1.5.2 Summary for Reptiles Freshwater turtles carry out important functions in freshwater ecosystems and sometimes play a keystone role in their environment. Freshwater turtle species were of concern during the SUNOCO oil spill that affected the John Heinz National Wildlife Refuge in suburban Philadelphia, and during the Enbridge Line 6b spill that affected the Kalamazoo River near Marshall, Michigan. Studies at the SUNOCO spill on the effect of crude oil exposure on female snapping turtle and painted turtle fertility, reproductive output, and development of offspring demonstrated that maternal oil exposure did not affect the adult stage through fertility or relative clutch mass, but it did affect the embryonic stages. Studies of the survival and behavior of freshwater turtles after rehabilitation following the SUNOCO spill concluded that rehabilitation of oil exposed freshwater turtles is effective in restoring these animals to normal behavior in nature. A total of 5600 oiled turtles, comprising eight species, were recovered following the Enbridge Line 6b spill, 98% of which were cleaned, rehabilitated and released. Turtle capture data suggest that turtles rehabilitated after an oil spill can be successfully reintroduced back into the spill area and survive.

6.1.6 Freshwater and Terrestrial Birds This section summarizes information available in scientific literature and government and industry reports on the recovery of freshwater and terrestrial birds following an oil spill. In the context of this summary, the measurable parameter for assessing recovery was dependent on the kinds of information available. Therefore, recovery can pertain to the return of a bird’s presence, relative abundance, density or population, behaviour (e.g., feeding success, habitat use) or community composition to pre-spill conditions in the absence of other potentially confounding factors. Recovery of freshwater and terrestrial birds after an oil spill will largely be dependent on: • extent of the effect (i.e., magnitude of the spill, number of birds affected), • complexity of the ecosystem affected (e.g., grassland, upland forest, riparian zones, bogs),

Page 6-30 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

• timeliness and thoroughness of clean-up actions, and • species’ vulnerability and resilience. Collectively, there is very limited science-based information on long-term effects and recovery of terrestrial and freshwater birds (Timoney and Lee 2009).

6.1.6.1 Case Studies for Recovery of Freshwater and Terrestrial Birds Freshwater birds are typically more vulnerable to oil spills than terrestrial birds due to their dependency on freshwater habitats for all or some of their life requirements and the likelihood that spilled oil will eventually be transported into freshwater systems. As well, the location and time of year in which an oil spill occurs can greatly influence the number of individuals and species affected. For example, a spill during the breeding season could negatively affect reproductive success for the year in addition to the removal of adults from the population. Additionally, a spill in a staging area used for foraging by birds during migration would also have a larger effect on populations compared to other times of the year. The following, in chronological order based on date of the incident, is a summary of effects, and subsequent recovery from, an oil spill on freshwater and terrestrial birds: • Experimental Study, Louisiana – Bird usage at three freshwater ponds subjected to an oil spill was found to be lower during the first 6-months after contamination compared to control ponds that were not subjected to oil (Chabreck, 1973). Within the oil-affected ponds, no dead birds were detected in the six months following contamination, and after that period bird use of those ponds increased (Chabreck, 1973). • , – In 1990, a spill of 31,500 barrels of oil in Arthur Kill – a narrow waterway that separates from New Jersey, USA – had immediate and long- term effects on a wading bird colony. It reduced reproductive success, changed foraging habitat quality and food availability, and may have caused birds to change foraging locations. Observations made the year before the spill (1989) and compared to observations in 1998- 1999 (10-11 years later) showed that flight patterns, foraging locations, and feeding success had returned to patterns documented before the spill for snowy egret (Egretta thula), great egret (Ardea alba) and black-crowned night heron (Nycticorax nycticorax) (Maccarone and Brzorad 2000). Reproductive success specifically had returned to pre-spill conditions within just 2-3 years (Parsons 1996). For glossy ibis, (Plegadis falcinellus), changes in foraging location after the spill had not returned to pre-spill conditions within 10 years. However, Maccarone and Brzorad (2000) note that a tripling in the number of glossy ibis nests since the spill, and subsequent changes in foraging location, may have been influenced by increased competition for food.

July 2012 Page 6-31

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

• Pine River, British Columbia – August 2000: 6,300 barrels of oil were released into a freshwater ecosystem (Table 6.4). Two birds affected by oil were captured – a golden eagle and a hooded merganser. The eagle was successfully released a few days after capture; the merganser died. At the time of the incident, BC MELP (2000) stated that a long-term assessment of the effects of the oil spill on fish and wildlife and their habitats, including areas that extend beyond the immediate vicinity of the river, was underway. No information pertaining to birds is presently available. • Wabamun Lake, Alberta – August 2005: 8,200 barrels of oil were released into a freshwater ecosystem as a result of a train derailment (Table 6.8). Notable effects to birds include the mortality of 300 western grebes. Annual counts of adult western grebes (based on number of nests multiplied by two) conducted at Wabamun Lake from 2001 to 2009 (Wollis and Stratmoen 2010) provide information on pre- and post-spill population levels. Prior to the spill (2001-2005), counts of adult western grebes ranged from 475 in 2005 to 1,500 in 2002 (Wollis and Stratmoen 2010). After the spill (2006-2009), counts ranged from 340 in 2009 to 1,100 in 2006 (Wollis and Stratmoen 2010). Although the mean population size of western grebe at Wabamun Lake was lower on average during the four years following the spill than in the four years prior to the spill, much of the decline in abundance occurred in 2004 and 2005, before the spill occurred. Human development and disturbance is a demonstrated factor contributing to the population decline of western grebes at Wabamun Lake (Ericson 2010). As well, population patterns from other lakes in the same region and for the same period, but without an oil spill, were similar to Wabamun Lake (Ericson 2010). • , British Columbia – May 2009: 1,300 barrels of oil released from a tank into a terrestrial ecosystem. No wildlife was reported to have been affected (Wyntonyk, 2009). No subsequent reports on effects or recovery are available. • Fort Greely, Alaska – May 2010: 1,900 barrels of oil released from a pump station into a terrestrial ecosystem. No indication that wildlife was affected or that clean-up efforts were delayed (Holland and Bluemink, 2010). No subsequent reports on effects or recovery are available. • Kalamazoo River, Michigan – July 2010: 18,700 barrels of oil released into a freshwater ecosystem resulting in 56 km of river closed for more than 14 months (Table 6.2). From July 2010 to July 2011, 196 birds were collected, of which 144 (73.5%) were released back into the wild (USFWS 2012). There are no reports on overall effect and subsequent recovery of bird species or communities. • Rainbow Pipeline, near Peace River, Alberta – April 2011: 28,000 barrels of oil released into both freshwater and terrestrial ecosystems (Table 6.9). The number of birds collected during initial clean-up included 21 ducks and 13 other migratory birds (Plains Midstream Canada 2012a). Clean-up has been ongoing for 14 months, but no additional reports of effects on birds, including the recovery of species or communities, are available. • Wrigley, Northwest Territories – May 2011: 1,500 barrels of oil released into terrestrial and freshwater habitats. No information on effects to birds available.

Page 6-32 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

• Red Deer River, Alberta – June 2012: 3,000 barrels of oil released into both freshwater and terrestrial ecosystems. As of 17 June 2012, 2 birds had died (Plains Midstream Canada 2012b). Clean-up activity is ongoing and it is too early to determine whether effects of the spill, and subsequent recovery, will be monitored. • Near Elk Point, Alberta – June 2012: 1,500 barrels of oil released at a pump station (Enbridge Inc. 2012) on agricultural land. No reports of wildlife effects.

Table 6.8 Summary Table for the Lake Wabamun Oil Spill Oil Spill Name Lake Wabamun Location Alberta, Canada Year 2005 Oil Type Bunker C Fuel Specific Name none Volume (metric tonnes) 126 Platform Train Environments Affected Freshwater Valued Ecological Component (VEC) Studied • Fish • Sediment • Water quality Summary: On August 3, 2005 a large train (43 rail cars) derailed and caused 126 metric tonnes of fuel (including Bunker C and pole treating oil) to spill on the north shore of the lake near the town of Whitewood Sands, Alberta. The oil spread to pelagic portion of the lake and continued to spread the north, south-east, and eastern shorelines. The massive cleanup ended in October of 2005 when the lake froze. The estimated volume of recovery from the initial cleanup efforts could not be found in the literature. Studies of a fish, sediment, and water quality recovery status were followed 5 years after the spill.

Table 6.9 Summary Table for the Plains Rainbow Pipeline Oil Spill Oil Spill Name Plains Rainbow Pipeline Location Alberta, Canada Year 2011 Oil Type Crude Specific Name none Volume (metric tonnes) 3,818 Platform Pipeline Environments Affected Terrestrial Summary: On April, 2011 the Rainbow pipeline ruptured and leaked 3,818 metric tonnes of crude oil. The leak occurred approximately 20 km southeast of Rainbow Lake in northwest Alberta. No detailed reports on cleanup efforts with volume recovered could be found in the literature.

July 2012 Page 6-33

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

6.1.6.2 Summary for Freshwater and Terrestrial Birds Despite the occurrence of oil spills in western Canada and elsewhere, few scientific studies document long-term effects and recovery to terrestrial and freshwater avian communities. However, based on the limited information that is available, it appears that birds associated with freshwater habitats are more vulnerable to an oil spill than species that use terrestrial habitats, and that terrestrial species that are ground-dwelling are more vulnerable to an oil spill than those species that live in trees. The speed and thoroughness of clean-up activities seems to reduce immediate and direct effects on birds, which likely serves to aid in long-term recovery. At least three studies (Chabreck, 1973; Maccarone and Brzorad, 2000; Ericson, 2010) have indicated that after being subjected to an oil spill, birds have recovered, or have remained relatively stable, relative to pre-spill conditions. Recovery periods have been in the order of 6 months to 10 years.

6.1.7 Terrestrial Wildlife and Wildlife Habitat The effects of an oil spill on terrestrial wildlife populations and their recovery is largely dependent on the extent of vegetation damage, time of year, and speed of vegetation regeneration (See Soil and Vegetation Section 6.1.1). British Columbia and portions of Alberta host habitat for grizzly bears, black bears (as well as the genetic variant of black bears, the white spirit bears), and wolves. Since there are no studies available on coastal wolves, in regards to health effects and recovery from oil spills, the focus of this section is on bears. There is also information on recovery concerning mule deer.

6.1.7.1 Grizzly Bears Grizzly bears are typically regarded as terrestrial predators that consume copious quantities of herbaceous vegetation, berries, insects, carcasses, and other mammals. However, coastal bears also rely on Pacific salmon, as the fish return to coastal streams to spawn (Hilderbrand et al. 1999b; Christensen et al. 2005). These same bears also can rely heavily on intertidal prey species, such as crabs, clams, barnacles, and mussels (Hatler et al. 2008), as well as intertidal vegetation, such as Lyngby’s sedge (Christensen 2008). Increased reliance on marine prey, such as fish, has a significant, positive effect on weight gain in individual bears (Robbins et al. 2007), and increased body mass is positively correlated to reproductive success and litter size (Hilderbrand et al. 1999a). The risks of damage from oil spills to large terrestrial predators, such as the grizzly, depend on several factors including the size and timing of the spill, toxicity of the oil, uptake into the grizzly bear food web, the level of exposure within the population, where the spill is located (terrestrial, marine, freshwater) and speed of clean-up (Sellers and Miller 1999). Studies on the effects of oil spills on grizzly bears have typically focused only on the marine coastal environment. Based on a diet consisting of heavy reliance on intertidal areas and marine prey species, bears could be susceptible to effects of oil spills, and be exposed to crude oil through a number of potential pathways. Coastal bears consume clay sediments directly to eliminate parasites from their gut. Bears also dig in soils and sediments to gain access to roots and whole plants (Tardiff and Stanford 1998; Christensen 2008) and clams (Hatler et al. 2008). As

Page 6-34 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments evidenced from other mammals, bears can be exposed to oil directly from consuming oil- contaminated carcasses or live prey, by grooming their own oil-contaminated fur, by inhaling hydrocarbon vapors from oil during the first few hours of a spill (which rapidly evaporate, see Section 2.0), or indirectly through consumption of prey (e.g., bivalves) that have assimilated hydrocarbons (Geraci and Williams 1990; Neff 1990; and Babcock and Short 1996 in Sellers and Miller 1999). Oil (and petroleum products) may even attract bears (BCWF 2011; GBSS 2012). Direct exposure can result in damage to eyes, digestive tract, liver, kidney, and nerves, along with hypothermia and behavioural abnormalities. The level of exposure to oil can vary with time of year of the oil spill because of seasonal habitat-use patterns. Despite the lack of investigations on exposure and effects of oil spills on large, terrestrial predators, oil exposure was documented in coastal grizzly bears following the Exxon Valdez oil spill in 1989 (Sellers and Miller 1999). In 1989, 42 million liters of crude oil was spilled from T/V Exxon Valdez, the oil covered 1,750 km of shoreline. A study conducted by Sellers and Miller (1999), both from Alaska Department of Fish and Game, examined the short- and long- term population dynamics of grizzly bears in the Katmai coast, approximately 500 km southwest of the spill site. Hydrocarbons were detected in 15% of the fecal samples collected in 1989 and 1990 indicating that hydrocarbons had been consumed by the bears. A dead yearling was found in a sedge meadow, which was found to have naphthalene and phenanthrene concentrations of 160 and 18 ppm, respectively, in the bile. The authors suggest this oil exposure either resulted in or at least contributed to the death of this bear (Sellers and Miller 1999). The authors also proposed that the disappearance of another yearling could have also been related to oil exposure; however, this conclusion is speculative as the yearling was never found. There was no difference in survival rates of female adult grizzly bears between oiled and unoiled areas (both rates were 96%) in 1989 to 1991 (short-term) or from 1992 to 1995 (92% and 90% survival, respectively; long-term). Survival of males was more difficult to quantify, in terms of percentage due to the temporary nature of the radio-collar, but there is no indication that oil was responsible for any deaths. There were also no significant differences between cub (approximately 36% for both oiled and unoiled areas) or yearling (46% and 77% for oiled and unoiled, respectively) survival from 1989 to 1995; however, sample sizes were low. The large difference in yearling survival was a result of the loss of three cubs from one mother, and it was not attributed to oil exposure. Recruitment (reproductive rate) and population growth rates were not significantly affected by the oil spill (Sellers and Miller 1999). The lack of effects was attributed to the timing of the spill – the spill reached the study areas (late April, early May) when the grizzly bears were still at higher elevations, when oiled carcasses would have been available in the lower elevation intertidal areas (Sellers and Miller 1999). Although polar bears do not occur on the Pacific coast of Canada, the results of an experiment with captive polar bears may be instructive (Oritsland et al. 1981). Three captive polar bears were exposed to crude oil that had been released onto the surface of an experimental pool. The bears ingested oil directly through grooming activities. Two of the bears died from erythropoietic dysfunction (i.e., decreased production of red blood cells) and renal abnormalities, while the

July 2012 Page 6-35

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

third bear became very ill but recovered after being removed from the oiled enclosure. Symptoms peaked at 5 to 6 weeks after initial exposure, with the third bear fully recovering after 5 months following exposure. This research established that ingestion of oil can be serious, although it should be noted that under the experimental conditions, the bears were not able to avoid exposure and the extrapolation of the results to the environment is problematic.

Bear Habitats Many of the potential effects on grizzly bears and the time it takes for populations to recover following a spill will be related to the extent of habitat loss, and the resulting decrease in food availability and prey species abundance. In terrestrial environments, the effects of an oil spill on wildlife populations and the recovery of these populations depends on the extent of habitat damage, time of year, and speed of vegetation regeneration. The section on Soil and Vegetation (Section 6.1.1) discusses recovery of these components. Recovery of soil and vegetation from an oil spill varied from a few months in a growing season to several decades. Woody plants and perennials may sustain little long-term damage from oil, but herbaceous plants and annuals may be severely affected (Baker 1971; Belsky 1975). Winter oiling of seeds probably reduces spring germination, and seedlings rarely recover after oiling; therefore, an oil spill in winter or spring may be most damaging to wildlife habitat. In the spring, omnivorous and herbivorous mammals, such as grizzly bears and moose, can rely heavily on herbaceous plants, such as Carex spp., as part of their diet. Although most herbaceous plants do not generally survive an oil spill, some species, such as Carex sp., can potentially regenerate in as few as two growing seasons after an oil spill, as observed in an alpine meadow near Mt. Baker, Washington (Belsky 1975). Nine years following the oil spill, Carex had re-established, covering 10 – 20% of the area (Belsky 1982). In mountainous regions with high snowpack, large volumes of snowmelt water may promote recovery by flushing much of the oil out of the area, which can lead to rapid recovery of habitat and improved seed germination within a year. In other regions, such as the tundra and boreal forest, plant community recovery may be slower (Collins et al. 1994). Diesel oil spills may affect vegetation differently from the way that they are affected by crude oil spills. Hutchinson (1984) suggests that diesel oil has more of a severe effect on ground vegetation than crude oil because toxic components persist in the soil for longer.

Coastal Habitats for Grizzly Bears Adverse effects to Pacific salmon survival, recruitment, development, or spawning activity significant could result in adverse effects to some populations of coastal grizzly bears. Geiger et al. (1996) estimated a loss of approximately 2.3 million wild pink salmon in Prince William Sound, Alaska, from 1990-1994, representing a small overall effect from the return of 144 million pinks. While this was considered an insignificant effect in the context of the whole of the Sound, it could represent an effect to a few local populations of bears if the loss of those 2.3 million salmon was from only a small number of streams. The Exxon Valdez Oil Spill Trustee Council (EVOSTC) concluded that the overall effect was an approximately 2% decrease in pink salmon from Prince William Sound, and that the salmon fishery did not fully recover commercially for 12 years (Section 5.1.5.2).

Page 6-36 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

6.1.7.2 Mule Deer Mule deer are primarily browsing herbivores that eat a variety of plants related to local conditions. Both sub-species that occur in the project area (Rocky Mountain mule deer and Sitka black-tailed deer) prefer browsing on Douglas-fir, Saskatoonberry, and willows, although they also eat many species of forbs and grasses (Shackleton 1999). Rocky Mountain mule deer seem to eat more grasses than do the Sitka black-tailed deer. During the summer, deer prefer habitat that is at or above the treeline, where a greater biomass of forage is available. In the fall and early winter, snow accumulation at the high elevations forces deer to lower valleys, where they browse on evergreen forbs and woody plants such as Vaccinium. Rocky Mountain mule deer will paw through the snow to eat the dead remains of large leafy forbs, such as cow-parsnip, that have turned into natural silage (Shackleton 1999). On the coast of Alaska in winter, when plants become scarce, Sitka black-tailed deer concentrate on beaches and increase their reliance on intertidal flora for food, such as kelp (Reynolds 1979). Deer are also known to swim between islands along the coast (Shackleton 1999). Similar to grizzly bears, deer may be susceptible to effects of oil spills in both marine coastal and terrestrial environments, either directly, through exposure to crude oil, or indirectly, through loss of habitat. Deer can be exposed directly by consuming oil-contaminated plants, by grooming their own oil-contaminated fur, or by inhaling hydrocarbon vapors from oil (during the first few hours after the spill), or indirectly by eating plants that have assimilated hydrocarbons from the environment. Deer may also be excluded from some areas soon after a spill by avoiding the disturbance created by clean-up operations. Deer use of oiled beaches and deer mortality was assessed in the year following the Exxon Valdez oil spill in Prince William Sound in 1989 (Lewis and Calkins 1995). Initially, few deer were observed on the oiled beaches, which may be a result of displacement due to human disturbance during the beach clean-up. However, in the following winter, deer continued to use beaches that had been oiled, especially when snow depth was high (Lewis and Calkins 1995). Hydrocarbons were found in muscle tissue of 3 out of 24 deer sampled from oiled beaches of Prince William Sound and Shuyak Island, of which only one also had elevated concentrations of hydrocarbons in the liver. However, accidental contamination may be attributed to two tissue samples, which led to inconclusive results. The primary cause of deer mortality appeared to be starvation, but not directly as a result of oiling; malnutrition and winter starvation was the ultimate cause of deer mortality. Lewis and Calkins (1995) noted in 30 deer examined, 28 had lungworm, which was not oil-related. The population of deer in Prince William Sound appeared to have been in poor health at the time of the study, which, coupled with a harsh winter in 1988/1989 and an oil spill, may have resulted in the high number of mortalities in 1989 and 1990. This study does not provide sufficient data to determine more long-term effects of an oil spill on the deer population.

July 2012 Page 6-37

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

6.1.7.3 Summary for Terrestrial Wildlife Overall, there have been very few documented studies on effects and recovery of terrestrial mammals, such as bears and herbivores, following oil spills, despite extensive knowledge of their heavy seasonal use of the intertidal areas and reliance on spawning salmon. The study that was conducted on wild grizzly bears following the Exxon Valdez spill found no measureable effects at the population level, although one yearling may have died has a direct result of oil exposure (Sellers and Miller 1999). The study that was conducted on Sitka black-tailed (mule) deer was inconclusive, due to possible tissue sample contamination and generally poor health of the deer population at that time, which owed to prevailing environmental conditions (Lewis and Calkins 1995).

6.2 Human Environments Associated with Freshwater and Terrestrial Environments The potential effects of oil spills from pipelines on the human environment and the time required for recovery are dependent on the volume, location, the nature of the resources affected; the extent, timing and location of traditional and non-traditional activities in the area; and the duration of clean-up and recovery. These effects are often minimal where the spill is geographically contained to land or a contained water body, but can become more substantial if the spilled oil enters a lake or river. Unlike marine spills, there has been little academic research or published literature on the effects of oil spills in terrestrial or freshwater environments on the human environment. The information that is available consists mainly of newspaper or other media accounts of the effects of some of the larger spills. As a result, the following discussion of spill effects and recovery may contain a combination of fact and opinion. However, for the human environment, there is a subjective component in defining recovery, so the perceptions of events by local residents are as important as the objective measure of recovery. In the literature, there is some information on six recent hydrocarbon spill events in terrestrial or freshwater environments. These include: • In April 2011, 28,000 barrels of oil (4,452 cubic metres) spilled from the Plains Rainbow Pipeline System at a location 95 km north of the Town of Peace River in northwest Alberta. The pipeline system transports crude oil from the central Mackenzie Valley and the Hay- Zama fields in northern Alberta to Edmonton. The spill affected about 8 hectares (20.5 acres) of public land and was the largest pipeline discharge in Alberta in 36 years. • In May 2011, there was a small spill of light sweet crude oil (1,500 barrels or 240 cubic metres) at a location south of Wrigley in the Northwest Territories from an Enbridge pipeline that runs from Normal Wells to Zama Alberta. • In July 2010, 20,082 barrels (3,193 cubic metres) of crude oil spilled into the Kalamazoo River near Marshall, Michigan, from the rupture of a 30 inch pipeline owned and operated by Enbridge Energy

Page 6-38 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

• In July 2007, about 1,500 barrels (240 cubic metres) of heavy synthetic crude oil was spilled at the Westridge Dock Transfer Line in Burnaby British Columbia as a result of damage caused by en equipment operator digging a trench for a new storm sewer. Part of the spill entered Burrard Inlet. • In August 2005, a derailment of CN freight train near the hamlet of Whitewood Sands on Lake Wabamun, Alberta, resulted in a spill of 4,375 barrels (696 cubic metres) of heavy oil and 555 barrels (88 cubic metres) of wood preservative. About 21% of the oil and 52% of the wood preservative eventually entered Lake Wabamun. • In August 2000, the rupture of the Plateau Pipeline, owned by Pembina Pipeline Corp. resulted in 6,200 barrels (986 cubic metres) of oil being spilled into the Pine River upstream of Chetwynd in British Columbia. It was the most costly spill in Canadian history at that time. For these spills, the key information sources on effects on the human environment and recovery are publicly available official incident report records, where available, as well as articles found in the press. There seem to have been no academic studies of the effects of oil spills in terrestrial or freshwater environments on the human environment. There is very little information on effects on the human environment from the Plains Rainbow Pipeline spill because it occurred in a relatively isolated area, was 300 metres away from flowing water and was easily contained (Calgary Herald, 2011). Similarly, there is little information on the Wrigley spill, as it occurred in a relatively isolated area, and permafrost in the ground above the pipeline served to prevent upward movement of the oil, and none of the spilled oil reached the nearby Willowlake River (Enbridge, 2011). It must be noted that spill effects and recovery of the human environment can only be described in general terms because each spill often has associated factors or issues which make each situation unique. For the case studies, the factors that make them unique include: • spills occurring at a previously contaminated site (Pine River spill as per Holdberg, 2005). • contaminants other than oil being spilled (Wabamun) • poor communications between response agencies and local populations (Lake Wabamun Residents Committee, 2007) • weather effects (high water during the Kalamazoo spills (National Transportation Safety Board, 2012) and forest fires hampering clean-up of the Plains Rainbow Pipeline spill • inadequate resources being available for containment and clean-up (Wabamun as per Goodman, 2006) These unique features can affect response efforts and thereby affect the duration and effectiveness of recovery efforts.

July 2012 Page 6-39

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

6.2.1 Human Safety One of the first effects of a terrestrial or freshwater oil spill on the human environment is the immediate concern for human safety, potentially resulting in immediate evacuation of nearby residents. In the case of the Lake Wabamun spill, local residents were evacuated 30 minutes after the spill, but were allowed back to their homes later that same day (Lake Wabamun Residents Committee, 2007). For the Westridge spill, 11 houses were directly sprayed by crude oil, other adjacent residences were also affected, and 250 residents voluntarily left their homes (Transportation Safety Board, 2007). A number of emergency workers and two members of the public were also directly sprayed with oil. A news story reported that 50 homes had to be evacuated and were housed in hotels for several nights (CBC News, 2007). Eventually most residents were allowed to return to the area after four days (Canadian Press, 2007), but residents of five houses were not allowed to return home for four months or more until clean-up operations were completed (Provincial Court of British Columbia, 2011). A section of the Barnett Highway was closed for several days after the spill. Although the oil spewed into the air for about 20 minutes and there was a strong odour, the B.C. Centre for Disease Control reported that air quality tests taken on the day of the spill showed that concentrations of crude oil components were less than found at an average industrial work site (Vancouver Sun, 2007). For the Kalamazoo spill, six houses were evacuated the day after the spill because of odour (National Transportation Safety Board, 2012). Four days after the spill an evacuation map was produced and evacuation orders were issued to about 50 homes. The evacuation order was lifted on August 12, 18 days after the spill occurred (National Transportation Safety Board, 2012). During this time, evacuees were housed with friends or relatives or at commercial accommodation facilities. In addition, eight days after the spill, Enbridge Inc. announced that it would buy the properties of homeowners who lived within 200 feet of either side of the river and had their houses up for sale prior to the spill and provide compensation on a case-by-case basis where the appraised values of houses dropped as a result of the spill, even if those houses were not on the market prior to the spill (Enbridge, 2012). Some people living adjacent to the Kalamazoo spill complained about having headaches and other maladies after the spill. Three months after the spill, the Michigan Department of Community Health conducted four community surveys along the affected waterways and found that 320 individuals had adverse health effects and tracked 41 calls placed to the poison center. These individuals typically complained of headache, nausea and respiratory effects that were consistent with the published literature on the health effects associated with crude oil exposure (National Transportation Safety Board, 2012). Both the Pine River and Plains Rainbow spill occurred in relatively unpopulated areas so there was no need for evacuation. However, in the case of the Plains Rainbow spill, classes at Little Buffalo School in nearby Cadotte Lake were suspended and 120 students sent home because odours were causing disorientation, headaches and nausea. However, according to the Energy Resources Conservation Board, the odours and health concerns were not caused by the leak (Global News Edmonton, 2011).

Page 6-40 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

In the case of Lake Wabamun, a compensation program was eventually provided for about 1,600 residents of the area for loss of use of the lake. Higher amounts were offered to people living closest to the lake, with the largest amounts being about $15,000 (CTV, 2008). The compensation agreement was negotiated three years after the spill occurred. Thus, available information suggests that the immediate effects of terrestrial and freshwater oil spills on human safety, including evacuation, typically lasted for about one week, although this took four months for the Burrard spills because some adjacent property were heavily oiled.

6.2.2 Drinking Water and Water Use A second major concern from a terrestrial or freshwater oil spill relates to potential effects on drinking water and other types of water use. When oil spills into water bodies that are being used for domestic or other water users, drinking water and other advisories are usually issued and sometimes restrictions are also placed on the use of groundwater. In the case of the Pine River spill, a drinking water advisory was issued on the same day the spill occurred (BC Ministry of Environment, 2000) and two days later the water supply to Chetwynd was cut-off over concerns about oil contamination of the water supply (BC Attorney General, 2000). Although the Town had sufficient water reserves for four to six weeks, drinking water was hauled into the community (Green, 2000). Two weeks after the spill, water was still being brought into Chetwynd, and the community had established plans to develop an alternate supply using groundwater (Green, 2000). Chetwynd is still primarily reliant on surface water from the Pine River but has a groundwater well as back-up. Parts of the $5 to $6 million in costs paid by the Pembina Pipeline Corp. for third part economic losses were associated with the disruption of the town’s water supply. Similar concerns arose as a result of the Kalamazoo spill. Within two days of the spill, the county health department had warned residents about use of water and set-up bottled water distribution centres. At the same time, the Michigan Department of Agriculture issued a ban on surface water withdrawals for crop or lawn irrigation or animal watering (National Transportation Safety Board, 2012). Four days after the spill, the county health department had also issued an advisory for people with private wells within 200 feet of the edge of the affected portions of the river. The ban on water use in the lowest reaches of the river was lifted 40 days after the spill but, for the other reaches of the river, the bottled water advisory lasted until November 6, 105 days after the spill occurred. For groundwater, Enbridge was ordered on September 23 to conduct sampling of all private and public water wells within 200 feet of the affected river and it (Section 6.1.2.1). Enbridge submitted its assessment on October 31, with the result that the drinking water ban was lifted and the bottled water program discontinued (National Transportation Safety Board, 2012). As of January 2012, Enbridge was continuing its water well testing program. At Lake Wabamun spill, there is limited use of lake water for drinking, but it is used for showering and other domestic purposes. One month after the spill, Capital Health (2005) advised residents closest to the spill area not to use well water for any purposes, due to possible contamination from the wood preservative. Capital Health also advised all residents not to use

July 2012 Page 6-41

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments lake water for drinking purposes but modified its lake water use advisory to allow water use for other purposes in some parts of the lake. The ban on use of lake water for drinking was still in place as of June 2006, nearly one year after the spill (CAW/TCA, 2006), but results of testing conducted in 2006 concluded that the remaining contamination no longer represented any threat to human health (Edmonton Journal, 2007). Thus, available information shows that terrestrial and freshwater oil spills can lead to restrictions being placed on drinking water from both surface and groundwater supplies. For the case studies, restrictions were in place for a month or for several years, during which time local residents were provided with alternate sources of drinking water. Similar restrictions were put on water used for agricultural or other purposes.

6.2.3 Food Quality Another issue associated with a freshwater oil spill on the human environment is the potential for tainting of fish and waterfowl and the resulting effects on food consumption. One month after the spill, Capital Health (2005) provided residents of Wabamun Lake with a letter that advised them against eating fish or waterfowl from the lake and that waterfowl with visible external oil should not be consumed. The letter noted that signs warning the public not to eat fish were posted around the lake. It also noted that birds that were oiled, cleaned and released had been banded with a specific set of numbers and that any bird with these numbers should not be consumed. Studies completed in 2006 concluded that remaining oil contamination posed no threat to human health but, in 2007, Capital Health was still advising people not to eat fish from the lake (Edmonton Journal, 2007). A fish and waterfowl consumption advisory for the lake still remained in place in 2008 (Sherwood Park News, 2008). While fishing on Lake Wabamun is currently allowed, all caught fish must be released as part of the fish recovery plans for the lake and not because of health concerns (Spencer, pers. comm.). For the Kalamazoo spill, there was still a ban on eating fish from the river two years after the spill occurred (Battle Creek Enquirer.com. 2012). Oil spills into freshwater can result in potential tainting of fish and waterfowl, with restrictions being placed on consumption. Such advisories are implemented immediately after the spill and, for the spills assessed in this study, lasted for two to three years after the spill.

6.2.4 Commercial Fisheries Effects of oil spill events on freshwater fish and other aquatic species can include direct mortality, some ingestion of oil and tainting, adverse effects on availability of food sources, and changes in reproduction. For the human environment, the most immediate effect of a spill is a ban on fishing (usually though controls on access to water bodies) and a ban on fish consumption. As observed for the Lake Wabamun, Kalamazoo and Pine River spills, these bans were put in place immediately after the spills occurred (BC Ministry of Environment, 2000b; National Transportation Safety Board, 2012).

Page 6-42 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

The resumption of normal fishing activities appears to be related to the recovery time for fish species. In the case of the Lake Wabamun spill, testing one year after the spill showed that, although whitefish eggs exposed to oil had 5% more deformities than normal, damage was expected to decrease over time and there will be no long-lasting damage (Edmonton Journal, 2007). However, a fish and waterfowl consumption advisory for the lake still remained in place in 2008 (Sherwood Park New, 2008). Lake Wabamun is currently being managed as a catch and release fishery in order to re-establish fish populations; the current prohibition on fish consumption is not related to health concerns (Spencer, pers. comm.). For the Kalamazoo spill, there are indirect restrictions on fishing because access to the river for boating is still restricted two years after the spill (mlive, 2012). The ban on eating fish from the river was to be lifted on June 21, 2012 (Battle Creek Enquirer.com. 2012). Overall, the available information shows that oil spills into freshwater result in restrictions being placed on fishing, with resumption of this activity depending on recovery of affected fish population and health concerns related to fish consumption. For the case studies, restrictions on fishing lasted one or two years, but recovery can be compromised where the oil spill is just one of several factors affecting fish populations.

6.2.5 Recreational Activities Oil spills can also adversely affect other types of recreational activities in freshwater environments. Oil spills have the ability to affect beach use, boating and swimming. The Lake Wabamun spill affected 12 kilometres of beach and parts of the lake were closed to swimming and boating. These closures remained in place for nearly one year, with Alberta Environment and the Capital Health Authority stating the lake was then safe for boating and swimming. However, two years after the spill, the Capital Health Authority was warning boaters and swimmers to avoid the oil patches and tar balls that remain in the lake, as contact can lead to skin irritation (CBC News 2007). Similar warnings were still being made five years after the spill (Global News, 2010). Ten days after the Pine Rivers spill, an information bulletin was issued asking residents to stay off the river between the spill location and Chetwynd until further notice (BC Ministry of Environment, 2000b). There is no information regarding the duration of closure on river boating and other activities. Two days after the Kalamazoo spill, a public health advisory was issued that asked people to avoid contact recreation and a ban on swimming and boating on the affected portion of the river was issued nine days after the spill (National Transportation Safety Board, 2012). Two years later, boating and other water-based activities are only being allowed on some sections of the Kalamazoo River. In April 2012, a three-mile section of the river (only half of which was affected by the spill) was opened for boating (mlive. 2012) but most of the river from the rupture site to Morrow Lake remained off limits to recreational users until July 2012 when an additional 34 miles of the river were reopened for recreation (USEPA, 2012).

July 2012 Page 6-43

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Oil from the Westridge spill entered the city sewer system and drained into Burrard Inlet, eventually fouling boats and beaches in the village of Belcarra and resulting in the closure of Cates Park in North Vancouver (Vancouver Sun, 2007). Belcarra Beach was closed for about a day and a kayak tour company temporarily closed its offices and suspended operations as a result of the spill (Vancouver Sun, 2007). Some beaches were closed for four months after the spill and 15 kilometres of beaches eventually required remediation (Provincial Court of British Columbia, 2011). Available information shows that oil spills, especially where oil enters freshwater, affects recreation, resulting in closures to boating and other recreational activities. Resumption of these activities is tied to clean-up operations having been completed and the authorities being confident that any remaining oil no longer poses a threat to human safety. While this has taken up to two years in the case of the Kalamazoo River, cautions on contact with residual oil were still in place five years after the Wabamun Lake spill.

6.2.6 Traditional Use and Cultural Activities There is no reported information on the actual effects of terrestrial or freshwater oil spills on traditional resources or cultures. While there is some evidence that oil spills have affected traditional use, the exact nature of the damages or recovery time are not specified. For some spills, First Nations have received compensation for oil spill effects on traditional resources, but the details of the damages and the basis of the settlement are often not made available to the public. In the case of the Lake Wabamun, oil reached the shores of land reserves belonging to the Paul First Nation 30 minutes after the Band was were told about the derailment (CTV 2006). The Paul First Nation subsequently initiated lawsuits against CN Rail for $505 million, for $200 million from the Federal Government and for $70 million from Alberta for damages to lands and resources as well as the effect on the residents' way of life. CN ultimately paid $10 million to the Paul First Nation, which announced that the money would be used to implement the band’s business plan and be distributed among the band’s 1,200 members (CBC. 2008). While there is no information on the time taken for First Nation traditional activities to recover, the compensation for damages was provided three years after the spill occurred. For the Wrigley spill, residents of Wrigley expressed concern about the potential effect of the spill on human health and the health of wildlife that people hunt for food. Enbridge committed to providing a health expert at the First Nation's annual assembly to address these issues (CBC News, 2011). However, Enbridge’s offer of $5,000 to the Pehdzeh Ki First Nation to allow them to hire an expert to analyze its detailed response plan was considered by the Chief to be insufficient (CBC News, 2011). In the case of the Westridge spill, there were concerns that the oil would adversely affect efforts by the Tsleil-Waututh First Nation’s marine stewardship program to rehabilitate 15 clam beds on the North Shore (Vancouver Sun 2007).

Page 6-44 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Overall, it is likely that, for the case studies examined, the recovery of spill effects on traditional activities depends primarily on how quickly these activities can resume, when harvest levels return to normal, and there are no concerns about food safety or tainting. Recovery is also related to the amount of time required to negotiate a financial settlement for damages and loss of traditional activities. For Lake Wabamun, this aspect of recovery took three years, and may have been longer than the time required for traditional resources to have recovered.

6.2.7 Industrial or Other Land and Resource Uses Available literatures provides little information about the effects of terrestrial or freshwater oil spills on industrial or other land use activities. This is partly because most of the major spills have occurred in relatively isolated areas where there are few other industrial activities that could be affected. One exception was the Lake Wabamun spill. Water from the lake was normally withdrawn for thermal power generation at two plants. On the day of the spill, the Trans-Alta Wabamun plant was shut down to prevent damage to plant equipment and further reduce environmental harm to the lake (TransAlta, 2005). TranAlta staff members were instrumental in providing first response containment for the spill. A phased restart of the Wabamun plant commenced on August 26, 2005, with TransAlta had to install screens, booms and various systems to capture oil to protect plant equipment and prevent the oil from recirculating into the lake. Full operation of the Wabamun plant resumed by September 11, 2005 (TransAlta, 2005), about 40 days after the spill occurred. It is understood that TransAlta received compensation from CN Rail for effects resulting from the spill. In the case of the Westridge spill, the clean-up operations resulted in closure of parts of Burrard Inlet for shipping, and this adversely affected the operations of Shell Canada. In 2007, Shell filed a lawsuit to recover costs associated with closing its operations for 11 days (Canada.com, 2008).

6.2.8 Effects of Clean-up Another potential effect of an oil spill on the human environment relates to the potential effects arising from clean-up activities. These activities typically involve people and equipment being brought into the affected areas, resulting in increased vehicle traffic on local roads, increased demand for commercial accommodation in nearby communities, and some employment opportunities for local residents. In some situations, the effects of clean-up operations on the human environment can be as disrupting as the spill itself, especially if the spill occurred in populated areas, and clean-up activities occur over an extend period of time. While there is little information on the effects of clean-up of terrestrial or freshwater spills on the human environment of nearby communities, these effects can last for up to two years. Clean-up of the Pine River oil spill was completed in two months (Goldberg, 2005). For the Wabamun spill, the on-water clean-up lasted one month and most of the shoreline clean-up was completed three months after the spill (Goodman, 2006). Some residual clean-up continued the following spring. Oil recovery from the Plains Rainbow spill occurred over three months, but site restoration is still underway 13 months after the spill occurred (Plains Midstream Canada, 2012).

July 2012 Page 6-45

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

Two years after the spill, clean-up of the Kalamazoo spill is still underway at the present time (July 2012), under the direction of the U.S. EPA and Michigan Department of Environmental Quality (Enbridge Inc., 2012b). There is minimal discussion of the extent to which clean-up of terrestrial or freshwater oil spills has provided employment for local residents. In all cases, professional companies are bought in to perform clean-up operations and there are sometimes opportunities for local residents. In the case of the Wrigley spill, the Chief of the Pehdzeh Ki First Nation Chief criticized Enbridge for failing to use local workers (CBC News, 2011). However, in operations such as the Trans-Alaska Pipeline, local aboriginal people have been trained as members of the spill response team. Local residents are often used in a volunteer capacity to clean waterfowl and other wildlife that had been oiled as a result of the spill. This was an important activity for both the Wabamun and Kalamazoo spills. On the Pine River spill, no one was allowed on the river to rescue any wildlife until four days after the spill so there were limited opportunities for animal rescue, with only two birds being brought into the Hillspring Wildlife Rehabilitation Facility in Dawson Creek (Green, 2000). For the Kalamazoo spill, Enbridge established a Wildlife Response Center in Marshall which cared for and released about 3,970 animals, including about 3,650 reptiles and 196 birds (National Transportation Safety Board, 2012). Over 5000 turtles were captured and the oiled turtles (almost 5000) were cleaned, rehabilitated and subsequently released (Robert Doherty, pers. comm.). Survival during capture and rehabilitation was about 97%. At the Wabamun spill, CN established a wildlife rehabilitation centre which received 18 species of birds and other wildlife (e.g., muskrats), including 622 birds in the 11 days after the event (Meredith, 2005). About 1000 volunteers signed up to help with the effort (Lake Wabamun Residents Committee, 2007).

6.2.9 Compensation and Litigation For some of the spills, companies had to pay compensation for third party damages. Such costs are not always publicly reported because they are directly negotiated between the company responsible for the spill and the third parties. Information on third party costs was available for only two of the five spills, and ranged from $5 to $45 per spilled. These costs were much higher in populated areas (like the Wabamun spill) and lower where there is low population density or where there are no other land or resource uses that were directly affected (Pine River). In most cases, it appears that compensation for damages was negotiated among the affected parties without need for the assistance of the courts and was completed fairly quickly. For example, the settlement for damages for the Wabamun spill was negotiated within three years of the spill. However, in the case of the Burrard Spill, one of the families directly affected by the spill filed a lawsuit for damages in 2010, three years after the spill actually occurred (sqwalk.com, 2010) and it is unclear as to when or if this will be resolved.

Page 6-46 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 6: Terrestrial and Freshwater Environments

6.2.9.1 Summary for the Human Environments Associated Terrestrial and Freshwater Environments Overall, past experience with terrestrial and freshwater oil spills from pipelines suggests that major incidents occur very infrequently. A review of the available literature for six recent spills shows: • The immediate effects of spills on human safety, including evacuation, can last a week, although this has taken as long as four months where property has been heavily oiled as a result of the spill (Burrard Inlet). • Another immediate effect can include restrictions on drinking water supplies. Such restrictions can be in place for a month or for several years, during which time local residents must be provided with alternate sources of drinking water. Similar restrictions may be put on water used for agricultural or other purposes. • Other immediate effects can include closures to fishing, closures to boating and other recreation, and restrictions on traditional harvesting. These can last for as short as time as two or three months, or until such time as clean-up operations have been completed and the authorities are confident that any remaining oil no longer poses a threat to human health and safety. Sections of the Kalamazoo River were only just reopened to water-based recreation two years after the spill and residents of Wabamun Lake were still being warned to stay away from any residual oil five years after the spill. • Clean-up activities have usually been completed within a couple of months but can extend to one to several years, especially if clean-up activities are halted during the winter months. • Other longer term effects may relate to potential tainting of fish and waterfowl and restrictions on consumption. Such advisories are implemented immediately after a spill and can last for two to three years after the spill. Overall, it appears that these aspects of the human environment recover from terrestrial and freshwater oil spills within two to four years following a major spill. While the recovery of the human environment from the physical effects of a spill can take up to four years, outstanding claims for compensation and damages may take longer to resolve. If the party responsible for the spill can successfully negotiate damage claims with affected parties, this can be a relatively short process. However, if parties fail to agree, resolution may require use of the courts and this process can take longer.

July 2012 Page 6-47

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 7: Discussion

7 Discussion The main question in this review is whether ecosystems and the human environments associated with them do or do not recover from oil spills; the answer is that recovery occurs. The scientific literature reviewed here is for cold temperate and sub-arctic regions and for marine, terrestrial, and freshwater environments. By far, recovery from marine oil spills has received greater study, and the scientific literature for marine spills is more voluminous and more readily available than for freshwater and terrestrial spills. Where recovery was followed, marine spills involved much greater volumes of spilled oil than freshwater and terrestrial spills (Appendices A and B). However, the bulk of the evidence is that for marine, freshwater and terrestrial oil spills, recovery of both the biophysical and human environments does occur.

7.1 Recovery Occurrence in Biophysical Environments The literature clearly reveals that recovery is not a rare occurrence; in fact, recovery is more common than not (Sections 5 and 6, Appendices A and B). For biophysical environments, there were 140 valued ecosystem components for which there were sufficient data for study authors to draw conclusions about recovery (Appendices A and B, Figure 7.1). The VECs associated with marine environments were 69% of this total, and 86% of these marine VECs were recovered or recovering at study’s end. VECs associated with the freshwater environment were 16% of the total and 70% of the freshwater VECs were recovered or recovering. The terrestrial VECs were 11% of the total and 50% of the terrestrial VECs were recovered or recovering. Of the total VECs examined from all environments, 80% were recovered or recovering. Where VECs had not recovered, this often reflected the study duration or complex interactions with other natural or man-caused factors (e.g., killer whale, herring). Another review of ecosystem recovery shows similar frequencies of recovery. In a review of the recovery of ecosystems, Jones and Schmitz (2009) examined 240 studies for evidence of ecosystem recovery from a broad range of human disturbances and events. For all types of disturbances, 34% of the indicator variables had recovered and 38% were recovering. For oil spills, they found that about 60% of the indicator variables examined showed recovery. For disturbances such as overfishing, logging, mining, eutrophication, and invasive species. recovery frequencies of indicator variables were less than 50%. Jones and Schmitz (2009) stated that “Our evidence does not support gloomy predictions.”

July 2012 Page 7-1

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 7: Discussion

Figure 7.1 The Number of VECs in Different Recovery Status, by Environment

Figure 7.2 The Number of VECs in Different Recovery Status, by Study Duration (years)

Page 7-2 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 7: Discussion

7.2 Recovery Time in Biophysical Environments Recovery does take time; how much time depends on the environment, the VEC, and other factors. The average time to recovered status for the biophysical VECs examined here was 2.3 years for freshwater environments and 5.8 years for marine (Appendices A and B). The frequency of recovered freshwater VECs is highest for cases less than one year and few freshwater VECs have a time to recovery greater than two years. The frequency of recovered marine VECs is highest (33%) between 5 and 10 years, and about 8% of the marine VECs have recovery times greater than 20 years (Figure A.4). The recovery times for terrestrial VECs are more complicated. Terrestrial VECs that recover appear to do so in about 2 years, but in some circumstances time to recovery for terrestrial VECs can take long time periods (Table A.5). Several of the case studies examined in this review were intentional experiments to study recovery or ones with little or no effective cleanup. Commonly, cleanup for terrestrial spills involves the removal of oiled soil and vegetation, replacement with clean soil, and replanting. Where clean activities have not been applied or applied ineffectively, some terrestrial VECs that were examined in this review were still recovering after 20 years (i.e., vegetation and soils in arctic regions) or showed no evidence of recovery (i.e., a contaminated groundwater aquifer). As been demonstrated through numerous in-situ and ex-situ remediation projects throughout North America, including British Columbia and Alberta, prompt containment of hydrocarbons, direct removal of contaminated soil, water and vegetation, and appropriate site restoration measures (e.g., replanting with native species) can result in rapid rehabilitation of these sites. Advances in in-situ remediation and phyto-remediation of contaminated soil are also proving to be promising approaches for rehabilitating contaminated sites without substantially disturbing the affected sites. Other reviews of recovery show similar times to recovery. After reviewing marine spills, Kingston (2002) concluded that “environmental recovery is relatively swift, being complete within 2-10 years.” In a review of 51 cases studies of recovery of coastal ecosystems from a variety of disturbances, Borja et al. (2010) concluded that recovery can occur in less than 5 years, but recovery from long-term insults, such as mine tailings and chronic wastewater discharges, may take decades. Borja et al. indicated that recovery of fish and macroinvertebrates from oil refinery discharges takes 2 to 3 years and recovery of intertidal and sub-tidal fauna from the same types of discharges takes 2 to 10 years. Jones and Schmitz (2009) reviewed a number of types of disturbances and ecosystems and found that time to recovery from oil spills was less than 5 years while time to recovery from overfishing was 10 to 20 years. Recovery of forested systems from logging and deforestation takes 3 to 4 decades, essentially the length of the cycle to regrow trees.

July 2012 Page 7-3

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 7: Discussion

7.3 Factors Influencing Recovery of Biophysical Environments A number of studies have demonstrated that recovery from marine spills is a function of the degree to which the system is sheltered from physical oceanographic processes (Clark 1982, Teal and Howarth 1984). At one extreme are open marine waters that recover rapidly in weeks or months, and at the other are sheltered, soft-sediment marshes that recover slowly over two or more decades (Teal and Howarth 1984). In the middle are headlands and exposed rocky shores that take 1 to 4 years to recover (Baker 1999). Because of the long recovery times for sheltered systems, modern spill response gives high priority to preventing oil from entering marshes and other similar systems. The same physical, chemical, and biological processes (e.g., spreading, dispersion, evaporation, biodegradation) that govern the behavior of oil in the marine environment also govern oil behavior in the freshwater environment (API 1999, NRC 1999). Where freshwater systems are fast flowing streams and rivers, turbulence can disperse oil into the water column and enhance dissolution and evaporation. Where freshwater systems are slow flowing or relatively still, oil can accumulate and persist longer Clean-up is generally a positive factor in ecological recovery from oil spills but it can be a negative in some instances. The two exceptions to the recovery time of 1 to 4 years for rocky shores were Esso Bernicia and Torrey Canyon for which recovery took at least 10 years (Baker 1999). These longer recovery times were associated with bull-dozing of the shore in the Esso Bernicia and the use of first generation dispersants in the Torrey Canyon. During the Amoco Cadiz, removal of oiled sediment from salt marshes with heavy earth-moving equipment lowered the level of the marshes and changed the patterns of sediment deposition and conditions for the growth of marsh plants. Modern spill response includes procedures to carefully select the most appropriate treatment for the oil type, level of contamination, and the nature of the shoreline (e.g., exposed and rocky versus sheltered and soft sediment). The characteristics of the spilled oil and the receiving environment interact to determine the environmental fate of the oil and its persistence in the environment (Section 2.0, 1999, 2003). Appropriate clean-up is a positive factor in ecological recovery by decreasing the level and duration of the oil in the environment. A consistent theme in Sections 5.1 and 6.1 points to two factors related to the life history of the organisms involved. First, recovery proceeds more rapidly when there is an abundant supply of propagules close to the affected area. Pelagic larvae in marine and freshwater environments enable recruitment from adjacent non-oiled areas. Second, a long life span means a long recovery time. This factors leads to many species of birds and mammals having longer recovery times than fish when the recovery depends on population growth by local reproduction rather than by immigration for other areas. Recovery status and recovery time also depend on how recovery is defined. This observation has been made in reviews of oil spills since the early 1980s through the present day (Teal and Howarth 1984; Kingston 2002; Jones and Schmitz 2009; Borja et al. 2010). These cited reviews all identify problems with defining recovery as a return to historical conditions. They

Page 7-4 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 7: Discussion

recommend that defining recovery as a return to a functioning state that provides valuable ecological goods and services is more appropriate.

7.4 Enhancing Recovery of Biophysical Environments There are cases where recovery may benefit from active human intervention beyond routine clean-up operations. The most protracted recovery appears to be in terrestrial environments where oil reaches groundwater (Section 6.2.1), where cold temperatures decrease biodegradation rates in soil, or when short thaw seasons limit vegetation growth (Section 6.1.2). The state of the art in bioremediation and phytoremediation continues to improve and would appear to be advantageous for spills where natural environmental processes for oil break down are inhibited.

7.5 Recovery in Humans Environments Recovery also occurs in the human environments associated with marine, freshwater, and terrestrial environments. Although recovery in the human environments depends on the completion of clean-up activities and recovery of harvested resources such as fish, shellfish, and wildlife, human environments have additional dynamics that shape their recovery. Immediate effects on the human environment derive from the clean-up activities, as well as safety closures and bans for harvesting and recreational areas. Recovery from these immediate effects can take from months to a few years. Examination of the course of recovery following the Exxon Valdez and the Selendang Ayu spills reveals several factors in recovery of human environments (Section 5.2). In the Exxon Valdez spill, protracted litigation and delayed compensation prevented recovery of the human environment for two decades. In contrast, following the Selendang Ayu spill, the settlement process resolved most claims within 5 years and without protracted litigation. Also, in the same spill, a modern Incident Command structure that took human factors into account, and modern spill response capabilities implemented by a dedicated response team were credited with earning the trust of the local residents and reducing effects of clean-up activities on the human environment to less than 2 years.

July 2012 Page 7-5

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 8: Conclusions

8 Conclusions Recovery after oil spills does occur. A review of the scientific literature does not support statements made in Intervener Evidence and oral testimony to the effect that the biophysical and human environments do not recover from oil spills. Recovery is common. The literature is clear that recovery from oil spills is not rare but rather occurs commonly. Recovery from marine spills has received greater study than that from freshwater and terrestrial spills. This review examined 50 spills and 174 Valued Ecosystem components (VECs) from the biophysical environments of cold temperate and sub-arctic regions. Study authors concluded that the VECs achieved a recovered state or recovering state by the study’s end in 86%, 70%, and 50% of the cases for marine, freshwater, and terrestrial biophysical environments, respectively. Another study found 60% of the indicator variable recovered by the study’s end after oil spills in several regions of the world. Where VECs had not recovered, this often reflected the study duration, or complex interactions with other natural or man-caused factors (e.g., killer whale, herring). Recovery takes time. The average time to recovered status for the biophysical VECs examined here was 2.3 and 5.8 years for freshwater and marine environments, respectively. The recovery times for terrestrial VECs are more complicated. Terrestrial VECs that recover appear to do so in about 2 years, but some terrestrial VECs examined here were still recovering after 20 years, or showed no recovery, often reflecting ongoing contamination (e.g., groundwater) and/or lack of appropriate clean-up. Several other reviews have indicated that recovery of marine environments from oil spills takes 2 to 10 years. For human environments, recovery appears to take from 2 to 5 years unless there is protracted litigation. Engagement of communities in determining spill response priorities and developing community mitigation plans can greatly aid in reducing effects and speeding recovery. The time to recovery depends on the environment, VEC, and other factors. Exposed rocky environments recover within a few years whereas sheltered, soft sediment environments such as marshes, can take two decades to recover. Similarly, fast moving freshwater systems tend to recover more quickly than slow flowing freshwater systems. VECs with short life spans can recover within days to a few years but those such as some birds and mammals can take longer. Several factors impede recovery. First, if oil persists, recovery can be slowed. Persistence depends on the characteristics of the oil and the environment. Clean-up is undertaken to remove and lessen the persistence of hydrocarbons. Second, inappropriate clean-up techniques, such as the use of heavy equipment in marshes, can substantially increase recovery time for biophysical environments. Fortunately, modern spill response is based on lessons learned and does seeks the most appropriate response given the type of environment oiled and other conditions. Third, protracted litigation and delayed compensation can delay recovery of human environments substantially. In recent spills, settlement processes without protracted litigation have led to resolution of claims in within 5 years.

July 2012 Page 8-1

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 8: Conclusions

In some cases, recovery needs active human aid. In terrestrial environments where cold temperature decrease biodegradation rates in soil and short thaw seasons limit vegetation growth, active remediation and restoration may be needed to enhance recovery. Fortunately, the state of the art in bioremediation and phytoremediation continues to improve and would appear to be advantageous for spills where natural environmental processes for oil break down are inhibited.

Page 8-2 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

9 References Agler BA, Kendall SJ, Irons DB, Klosiewski SP. 1999. Declines in marine bird populations in Prince William Sound, Alaska coincide with a climactic regime shift. Waterbirds 22: 98-103. Alaska Department of Fish and Game. 2005. ADF&G Announces Reopening of Commercial Fishing in Skan and Makushin Bays. Available at: http://dec.alaska.gov/SPAR/PERP/RESPONSE/SUM FY05/041207201/041207201 index.htm Alaska Regional Response Team (ARRT) 2002. Alaska Implementation Guidelines for Federal On- Scene Coordinators for the Programmatic Agreement on Protection of Historic Properties during Emergency Response under the National Oil and Hazardous Substances Pollution Contingency Plan. Anchorage, AK, USA: U.S. Department of the Interior, Office of Environmental Policy and Compliance. Alberta Sustainable Resource Development and Alberta Conservation Association (ASRD and ACA). 2006. Status of western grebe in Alberta. Alberta Sustainable Resource Development, Wildlife Status Report No. 60, Edmonton, AB. 29 pp. Allan, J.D. 1995. Stream Ecology: Structure and function of running waters. 1st Edition. Published by Chapman & Hall. AMAP 2008. Oil and Gas Activities in the Arctic: Effects and Potential Effects. Arctic Monitoring and Assessment Programme, Oslo, Anchorage Daily News. 1999a. Spill funds are not end of trouble. Available at: http://www.adn.com/evos/stories/T99032619.html Anchorage Daily News. 1999b. Years later spill’s toll still rising. Available at: http://www.adn.com/evos/stories/T99032267.html Anchorage Daily News. 2010. Criminal charges were leveled after Exxon Valdez spill. Available at: http://www.adn.com/2010/06/01/1303517/criminal-charges-leveled-after.html#storylink+cpy Anderson, D. W., Newman, S. H., Kelly, P. R., Herzog, S. K., and K. P. Lewis. 2000. An experimental soft-release of oil-spill rehabilitated American coots: 1. Lingering effects on survival. Environmental Pollution 107: 285-294. http://www.sciencedirect.com/science/article/pii/S0269749199001803. Anderson, Ed. 2011. Oil spill impact on Tourism through 2013. Available at: http://www.gnohla.com/latest-news/oil-spill-impact-on-tourism-through-2013.html Angermeier, P.L., and J.R. Karr. 1986. Applying an Index of Biotic Integrity Based on Stream-Fish Communities: Considerations in Sampling and Interpretation. North American Journal of Fisheries Management. 6:418-129. Arnold, S. 2006. M/V Selendang Ayu Oil Spill Unalaska, Alaska. Public Health Evaluation of Subsistence Resources Collected During 2005. Final Report Alaska Department of Health and Social Services, Division of Health Department of Epidemiology, Anchorage.

July 2012 Page 9-1

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

AURIS (1994) Scientific Criteria to Optimise Oil Spill Clean-up Operations and Effort. A report by AURIS Environmental, Aberdeen. 56pp plus appendices, figures and tables. Baca, B.J., Lankford, T.E., Gundlach, E.R. (1987) Recovery of Brittany coastal marshes in the eight years following the Amoco Cadiz incident. Proceedings of the 1987 International Oil Spill Conference. Baedecker, M.J., I.M. Cozzarelli, R.P. Eganhouse, D.I. Siegel, P.C. Bennett. 1993. Crude oil in a shallow sand and gravel aquifer: III. Biogeochemical reactions and mass balance modeling in anoxic groundwater. Applied Geochemistry 8, 569– 586. Baker JM, Guzman L, Bartlett PD, Little DI and Wilson CM (1993) Long-term fate and effects of untreated thick oil deposits on saltmarshes. In: Proceedings of the 1993 International Oil Spill Conference. Baker, J., Little, A. and Heaps, L. (1996) Guidelines for Assessing the Ecological Condition and Recovery of Oiled Shores. A report by AURIS Environmental, Aberdeen. 89pp. Baker, J.M. 1971. Seasonal effects of oil pollution on salt marsh vegetation. Oikos 22: 106-110. Baker, J.M., Clark, R.B., Kingston, P.F. and Jenkins, R.H. (1990). Natural Recovery of Cold Water Marine Environments After an Oil Spill. Presented at the Thirteenth Annual Arctic and Marine Oilspill Program Technical Seminar. pp 111. Baker, J.M., Guzman, L.M., Bartlett, P.D., Little, D.I., Wilson, C.M., 1993. Long-term fate and effects of untreated thick oil deposits on salt marshes. In: Proceedings of the 1993 International Oil Spill Conference. American Petroleum Institute, Washington, DC, pp. 395–399. Ballachey, B.E., J.L. Bodkin, S. Howlin, A.M. Doroff, and A.H. Rebar. 2003. Correlates to survival of juvenile sea otters on Prince William Sound, Alaska, 1992-1993. Canadian Journal of Zoology 81:1494-1510. Banks, M.K., Schwab P., Liu B, Kulakow, P.A., Smith J.S., Kim, R. 2003. The effects of plants on the degradation and toxicity of petroleum contaminants in soil: a field assessment. Advances in Biochemical Engineering and Biotechnology 78:75-96. Barbour, M.T., J. Gerritsen, B.D. Snyder, and J.B. Stribling. 1999. Rapid Bioassessment Protocols for Use in Streams and Wadeable Rivers: Periphyton, Benthic Macroinvertebrates, and Fish, Second Edition. EPA 841-B-99-002. U.S. Environmental Protection Agency; Office of Water; Washington, D.C. Barrett-Lennard, L.G. 2000. Population Structure and Mating Patterns of Killer Whales, Orcinus orca, as Revealed by DNA Analysis. PhD Dissertation, University of British Columbia, Vancouver, BC, Canada Batten S.D., Allen R.J.S., Wotton C.O.M., (1998), The effects of the Sea Empress oil spill on the plankton of the southern . Mar. Pollut. Bull. 36, 764-774. BC Ministry of Environment, Lands and Parks (BC MELP). 2000. Pine River oil spill response: Update #6. Information Bulletin, 11 August 2000. Available: http://www2.news.gov.bc.ca/archive/pre2001/2000/august/ib614.asp. Accessed: 17 June 2012.

Page 9-2 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Bell, B., J.R. Spotila, and J. Congdon. “High Incidence of Deformity in Aquatic Turtles in the John Heinz National Wildlife Refuge”. [In eng]. Environmental Pollution 142, no. 3 (Aug 2006): 457-65. Bell, B.A. 2005. The effects of crude oil contamination on the reproduction of freshwater turtles. PhD thesis. Drexel University, Philadelphia. Bell, S.A., Stevens, P.A., Norris, D.A., Radford, G.L., Gray, A.J., Rossall, M.J. and Wilson, D. 1999. Damage assessment survey of saltmarsh affected by the Sea Empress oil spillage. Report by Institute of Terrestrial Ecology to CCW. 45pp plus appendices and figs. Bell, S.A., Stevens, P.A., Norris, D.A., Radford, G.L., Gray, A.J., Rossall, M.J. and Wilson, D. (1999) Damage assessment survey of saltmarsh affected by the Sea Empress oil spillage. Report by Institute of Terrestrial Ecology to CCW. 45pp plus appendices and figs. Belsky, J. 1975. An oil spill in an alpine habitat. Northwest Science 49:141-146. Belsky, J. 1982. Diesel oil spill in a subalpine meadow: 9 years of recovery. Canadian Journal of Botany 60:906-910. Benson, A.J. and A.W. Trites. 2002. Ecological effects of regime shifts in the Bering Sea and eastern North Pacific Ocean. Fish and Fisheries. 3:95-113. Bliss, L.C. and R.W. Wein. 1972. Plant community responses to disturbances in the Western Canadian Arctic. Canadian Journal of Botany 50:1097-1109. Bodkin JL, Ballachey BE, Coletti HA, Esslinger GG, Kloecker KA, Rice SD, Reed JA, Monson DH. 2012. Long-term effects of the ‘Exxon Valdez’ oil spill: sea otter foraging in the intertidal as a pathway of exposure to lingering oil. Marine Ecology Progress Series 447: 273-287. Bodkin JL, Ballachey BE, Dean TA, Fukuyama AK, Jewett SC, McDonald L, Monson DH, O’Clair CE, VanBlaricom GR. 2002. Sea otter population status and the process of recovery from the 1989 ‘Exxon Valdez’ oil spill. Marine Ecology Progress Series 241: 237-253.Bodkin JL, Ballachey BE, Esslinger GG. 2011. Synthesis of nearshore recovery following the 1989 Exxon Valdez oil spill: Trends in sea otter population abundance in Western Prince William Sound. Exxon Valdez Oil Spill Restoration Project Final Report (Restoration Projects 070808, 070808A, and 090808), U. S. Geological Survey, Alaska Science Center, Anchorage, Alaska. Bodkin JL, Ballachey BE. 1997. Restoration notebook: sea otter (Enhydra lutris). Exxon Valdez Oil Spill Trustee Council (EVOSTC), Anchorage, AK. pp 1-8. Available: http://www.evostc.state.ak.us/Universal/Documents/Publications/RestorationNotebook/RN_seaot ter.pdf. Boehm PD, Neff JM, Page DS (2007) Assessment of Hydrocarbon exposure in the waters of Prince William Sound after the Exxon Valdez oil spill: 1989-2005. Mar Pollut Bull 54:339-367 Boehm PD, Page DS, Brown JS, Neff JM, Bragg JR, Atlas RM. Distribution and weathering of crude oil residues on shorelines 18 years after the Exxon Valdez spill. Environmental Science and Technology 2008; 42:9210–9216.

July 2012 Page 9-3

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Boehm PD, Page DS, Neff JM, Brown JS. 2011. Are sea otters being exposed to subsurface intertidal oil residues from the Exxon Valdez oil spill? Marine Pollution Bulletin 62: 581-589. Borja, A, DM Dauer, M Elliott, and CA Simenstad. 2010. Medium and long-term recovery of estuarine and coastal ecosystems: patterns, rates, and restoration effectiveness. Estuaries and Coasts 33:1249-1260. Bowman, T. 1999. Bald eagle (Haliaeetus leucocephalus), Recovery Notebook. US Fish and Wildlife Service. Prepared for Exxon Valdez Oil Spill Trustee Council, Anchorage, AK. Braddock, Joan F., Jon E. Lindstrom, and Roger C. Prince. “Weathering of a Subarctic Oil Spill over 25 Years: The Caribou-Poker Creeks Research Watershed Experiment”. Cold Regions Science and Technology 36, no. 1-3 (2003): 11-23. Braham et al. 1980 In frakers references Brannon, E. L., A. W. Maki, L. L. Moulton, and K. R. Parker. "Results from a Sixteen Year Study on the Effects of Oiling from the Exxon Valdez on Adult Pink Salmon Returns." [In eng]. Marine Pollution Bulletin 52, no. 8 (Aug 2006): 892-99. Brannon, E. L., L. L. Moulton, L. G. Gilbertson, A. W. Maki, and J. R. Skalski. An assessment of oil spill effects on pink salmon populations following the Exxon Valdez oil spill -Part 1: Early life history, pp. 548–584. In: Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters, ASTM STP 1219, (Wells, P.G., J. N. Butler, and J. S. Hughes, Eds.), Philadelphia, PA: American Society of Testing and Materials (1995). Brown, E.D., B.L. Norcross, and J.W. Short. 1996. Introduction to Studies on the Effects of the (Exxon Valdez) Oil Spill on Early Life History Stages of Pacific Herring, (Clupea pallasi), in Prince William Sound, Alaska. Canadian Journal of Fisheries and Aquatic Sciences 53: 2337-42. Brownell, R.L., Jr. 1971. Whales, dolphins, and oil pollution. In: Biological and oceanographic survey of the Santa Barbara Channel Oil Spill 1969-1970, D. Straughn, ed. Sea Grant Publ. No. 2, vol. 1, pp. 225-276. Allan Hancock Foundation, University of Southern California, Los Angeles, CA. Bue, B. G, S. Sharr, S. D. Moffitt, and A. K. Craig. Effects of the Exxon Valdez oil spill on pink salmon embryos and pre-emergent fry. Am. Fish. Soc. Symp. 18: 619–627 (1996). Bue, B.G., S. Sharr, and J.E. Seeb. 1998. Evidence of damage to pink salmon populations inhabiting Prince William Sound, Alaska, two generations after the Exxon Valdez oil spill. Transactions of the American Fisheries Society 127: 35-43. Bue, B.G., S. Sharr, S.D. Moffitt, and A.K. Craig. 1996. Effects of the Exxon Valdez oil spill on pink salmon embryos and pre-emergent fry. Proceedings, Exxon Valdez Oil Spill Symposium, Anchorage, Alaska, USA. February 2-5, 1993. Pp. 619 – 627. Burk, J.C. 1977. A four-year analysis of vegetation following an oil spill in a freshwater marsh. Journal of Applied Ecology 14:515-522. Burns, K.A., S.D. Garrity, and S. C. Levings. 1993. How many years until mangrove ecosystems recover from catastrophic spills? Marine Pollution Bulletin 26:239-248.

Page 9-4 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Bustard, D., and M. Miles. 2011. Potential effects of an oil pipeline rupture on reach 2 of Morice River. A submission to the Joint Review Panel – Enbridge Northern Gateway Project. 32 pp + Figures and Tables. California Department of Fish and Game. 2009. Pacific Herring Commercial Fishing Regulations. Schedule No. 98052052. Sections 163, 163.1, 163.5, and 164, Title 14. California Code of Regulations, California Department of Fish and Game. California Department of Fish and Game. 2010. Pacific Herring Commercial Fishing Regulations. Schedule No. 98052052. Sections 163, 163.1, 163.5, and 164, Title 14. California Code of Regulations, California Department of Fish and Game. California Department of Fish and Game. 2011. Pacific Herring Commercial Fishing Regulations. Schedule No. 98052052. Sections 163, 163.1, 163.5, and 164, Title 14. California Code of Regulations, California Department of Fish and Game. Carls M.G., S.D. Rice, J.E. Hose. 1999. Sensitivity of Fish Embryos to Weathered Crude Oil: Part I: Low-Level Exposure during Incubation Causes Malformations, Genetic Damage, and Mortality in Larval Pacific Herring (Clupea pallasi). Environmental Toxicology and Chemistry 18:491–493 Carls MG, Marty GD, Meyers TR et al (1998) Expression of viral hemorrhagic septicemia virus in prespawning Pacific herring (Clupea pallasi) exposed to weathered crude oil. Can J Fish Aquat Sci 55(10):2300-2309 Carls, M.G., G.D. Marty, and J.E. Hose. 2002. Synthesis of the Toxicological Impacts of the Exxon Valdez Oil Spill on Pacific Herring (Clupea pallasi) in Prince William Sound, Alaska, U.S.A. Canadian Journal of Fisheries and Aquatic Sciences 59: 153-72. Carson, Richard T. and W. Michael Hanemann. 1992. A Preliminary Economic Analysis of Recreational Fishing Losses Related to the Exxon Valdez Oil Spill. Available at: http://www.evostc.state.ak.us/facts/economic.cfm CBC News. 2011. Wrigley residents voice pipeline spill concerns. August 11, 2011. Available at: http://www.cbc.ca/news/canada/north/story/2011/08/11/nwt-wrigley-enbridge-meeting.html. Chabreck, R.H. 1973. Bird use of marsh ponds subjected to oil spills. Proceedings of the Louisiana Academy of Sciences 36: 101-136. Christensen, J.R. 2008. Factors affecting persistent organic pollutant (POP) accumulation in British Columbia grizzly bears (Ursus arctos horribilis). PhD Thesis. 2008. University of , Victoria, British Columbia. Christensen, J.R., M. MacDuffee, R.W. MacDonald, M. Whiticar, and P.S. Ross. 2005. Persistent organic pollutants in British Columbia grizzly bears: consequence of divergent diets. Environmental Science and Technology 39: 6952-6960. Clark, RB (ed) 1982. The Long-Term Effects of Oil Pollution on Marine Populations, Communities, and Ecosystems. The Royal Society. London.

July 2012 Page 9-5

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Clavero, M, L Bortons, P Pons, and D Sol. 2009. Preomoinent role of invasive species in avian biodiversity loss. Biological Conservation 142:2043-2049. Clifford, J. 2011. Alberta’s oil spill history. NICHE: Network in Canadian history and environment. Available: http://niche-canada.org/node/9992. Accessed: 17 June 2012. Coats, D.A., A.K. Fukuyama, J.R. Skalski, and S. Kimura. 1999. Monitoring of biological recovery of Prince William Sound intertidal sites impacted by the Exxon Valdez oil spill. 1997. Biological Monitoring Survey. NOAA Technical Memorandum NOS OR&R 1. National Oceanic and Atmospheric Administration, Seattle, WA. 73 pp. + appendices. Collins, C.M., C.H. Racine, and M.E. Walsh. 1994. The physical, chemical, and biological effects of crude oil spills after 15 years on a black spruce forest, interior Alaska. Arctic 47: 164-175. Conan, G. 1982. The long-term effects of the Amoco Cadiz oil spill. Phil. Trans. R. Soc. London B 297:323-33. Cooney RT, Allen JR, Bishop MA et al (2001) Ecosystem controls of juvenile pink salmon (Oncorhynchus gorbuscha) and Pacific herring (Clupea pallasi) populations in Prince William Sound, Alaska, fisheries. Fish Oceanogr 10(Suppl 1):1-13 Cooney, R. T., and T. M. Willette. Factors influencing the survival of pink salmon in Prince William Sound, Alaska, pp. 183–196. In: Estuarine and Ocean Survival, of Northeastern Pacific Salmon (Em- mett, R. L., and M. H. Schiewe, Eds.) NOAA Tech. Memo NMFS- NWFSC-29 (1997). Cooney, RT (1981) Bering sea zooplankton and microplankton communities with emphasis on annual production. In: Hood DW, Calder JS (eds) The eastern Bering Sea shelf: oceanography and resources, Vol 2, University of Washington Press, Seattle, pp 947-974 Costa DP, Kooyman GL. 1982. Oxygen consumption, thermoregulation, and the effect of fur oiling and washing on the sea otter, Enhydra lutris. Canadian Journal of Zoology 60: 2761-2767. Courchamp F., L. Berec, and J. Gascoigne. 2008. Allee Effects in Ecology and Conservation. Oxford University Press, Oxford, UK Crump, R.G., Williams, A.D. and Crothers, J.H. 2003. West Angle Bay: A case study. The fate of limpets. Field Studies, 10, 579-599. Crunkilton, R.L. and R.M. Duchrow. "Impact of a Massive Crude Oil Spill on the Invertebrate Fauna of a Missouri Ozark Stream." Environmental Pollution 63 (1990): 13 - 31. Cushman, Robert M., and Jonathan C. Goyert. "Effects of a Synthetic Crude Oil on Pond Benthic Insects." Environmental Pollution Series A, no. 33 (1984): 163-86. Dahlheim, M.E. 1988. Killer Whale (Orcinus orca) Depredation on Longline Catches of Sablefish (Anoplopoma fimbria) in Alaskan Waters. Processed Report 88-14, US Department of Commerce, National Marine Fisheries Service, Northwest and Alaska Fisheries Center, Seattle, WA, USA Dauvin, J-C. (1998) The fine sand Abra alba community of the Bay of Morlaix twenty years after the Amoco Cadiz oil spill. Marine Pollution Bulletin, 36 (9), 669-676.

Page 9-6 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Dauvin, J-C. (2000) The muddy fine sand Abra alba – Melinna palmata community of the Bay of Morlaix twenty years after the Amoco Cad0069z oil spill. Marine Pollution Bulletin, 40(6), 528-536. Davenport, J. (1982) Oil and planktonic ecosystems. In: Clark, R.B. (ed). The long-term effects of oil pollution on marine populations, communities and ecosystems. Phil. Trans. Royal Society, London, B, 297, 369-384. Davies, J.M., McIntosh, A.D., Stagg, R., Topping, G. and Rees, J. 1997. The fate of the Braer oil in the marine and terrestrial environments. In: The Impact of an Oil Spill in Turbulent Waters: The Braer, pp. 26–41 (Davies, J.M. and Topping, G., eds.). The Stationery Office Limited, Edinburgh, UK. ISBN 0-11-495798-3. Day, B. 2005. Ecological effects to benthic infauna from lingering oil 15 years after the Exxon Valdez oil spill. Restoration Project 040772. Prepared for the Exxon Valdez Oil Spill Trustee Council, Anchorage, AK. Integral Consulting Inc., Mercer Island, WA. Day, R.H., Murphy, S.M., Wiens, J.A., Haywards, G.D., Harner, E.J., Smith, L.N. 1997a. Effects of the Exxon Valdez Oil Spill on Habitat Use by Birds in Prince William Sound, Alaska. Ecological Applications. 7: 593-613. Day, R.H., Murphy, S.M., Wiens, J.A., Hayward, E.J., Lawhead, B.E. 1997b. Effects of the Exxon Valdez oil spill on habitat use by birds along the Kenai Peninsula, Alaska. The Condor. 99: 728- 742. Day, R.H., S.M. Murphy, J.A. Wiens and K.R. Parker. 2003. Changing habitat use by birds after the Exxon Valdez Oil Spill. International Oil Spill Conference Proceedings: Vancouver, B.C. de Pennart, H., R. Crowther, T. Taylor, M. Morden, and S. Mattison. "The Use of Ecological Risk Assessment for Regional Management of Aquatic Impacts." Environmental Services Association of Alberta Symposium (2004). Dean TA, Bodkin JL, Fukuyama AK, Jewett SC, Monson DH, O’Clair CE, VanBlaricom GR. 2002. Food limitation and the recovery of sea otters following the ‘Exxon Valdez’ oil spill. Marine Ecology Progress Series 241: 255-270. Dean TA., and Jewett SC. 2001. Habitat-specific recovery of shallow subtidal communities following the Exxon Valdez oil spill. Ecological Applications 11:1456–1471 Dean, T.A., Jewett, S.C., Laur, D.R. and Smith, R.O. (1996b) Injury to epibenthic invertebrates resulting from the Exxon Valdez oil spill. In: Proceedings of the Exxon Valdez Oil Spill Symposium, held in Anchorage, Alaska, February 1993. American Fisheries Society Symposium 18, 424-439. Dean, T.A., Stekoll, M.S. and Smith, R.O. (1996a) Kelps and oil: the effects of the Exxon Valdez oil spill on subtidal algae. In: Proceedings of the Exxon Valdez Oil Spill Symposium, held in Anchorage, Alaska, February 1993. American Fisheries Society Symposium 18, 412-423. Dean, T.A., Stekoll, M.S., Jewett, S.C., Smitha, R.O. and Hose, J.E. 1998. Eelgrass (Zostera marina L) in Prince William Sound, Alaska: effects of the Exxon Valdez oil spill. Marine Pollution Bulletin, 36: 201–210.

July 2012 Page 9-7

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Department of Health and social Services and U.S. Fish and wildlife Services. 2006. Fact Sheet, M/V Selendang Ayu Oil Spill, Subsistence Foods Consumption Safety. Available at: http://dec.alaska.gov/SPAR/PERP/RESPONSE/SUM FY05/041207201/041207201 index.htm Deriso, R.B., M.N. Maunder, and W.H. Pearson. 2008. Incorporating Covariates into Fisheries Stock Assessment Models with Application to Pacific Herring. Ecological Applications 18: 1270-86. Deshaies, A., Boudreau, S., Harper, K.A. 2009. Assisted Vegetation in a Subarctic Environment: Effects of Fertilization on the performance of Three Indigenous Plant Species. Arctic, Antarctic and Alpine Research, 41(4): 434-441. DFO. 2008. Population Assessment: Steller Sea Lion (Eumetopia jubatus). DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2008/047. Elmgren, R.; Hansson, S.; Larsson, U.; Sundelin, B. and Boehm, P.D. 1983. The "Tsesis" oil spill: acute and long-term impact on the benthos, Mar. Biol. (Berlin), 73 (1): 51-65. Elston, R.A., and T.R. Meyers. 2009. Effect of Viral Hemorrhagic Septicemia Virus on Pacific Herring in Prince William Sound, Alaska, from 1989 to 2005. Diseases of Aquatic Organisms 83: 223-46. Enbridge Inc. 2012. Overview of Marshall Spill. Available at: http://csr.enbridge.com/index.php/pipeline- integrity/marshall-spill-case-study Enbridge Inc. 2012b. Line 6B response. Available at: http://response.enbridgeus.com/response/main.aspx?id+15705 Enbridge Inc., 2012. Enbridge reports oil release at pump station facility. Available: http://www.enbridge.com/MediaCentre/News.aspx?yearTab=en2012&id=1630491. Accessed 20 June 2012. Enbridge. 2011. Overview of Norman Wells Spill. Available at: http://csr.enbridge.com/index.php/pipeline-integrity/norman-wells-spill-case-study Energy Resources Conservation Board (ERCB). 2009. Public Safety / Field Surveillance Provincial Summary 2008. ST57-2009. Available: http://www.ercb.ca/docs/products/STs/st57-2009.pdf. Accessed: 17 June 2012. Engelhard, G.H., and M Heino. 2006. Climate change and condition of herring (Clupea harengus) explain long-term trends in extent of skipped reproduction. Oecologia. 149:593-603. English.new.cn. 2011. Experts play down effects of oil disaster on New Zealand tourism. Available at: http://news.xinhuanet.com/english2010/world/2011-10/14/c 131192135.htm Environment News Services. 2009. 20 Years After Valdez Oil Spill, Exxon Still Owes $92M. Available at: http://www.ens-newswire.com/ens/mar2009/2009-03-24-01.asp Ericson, M.E., 2010,. Persistence and abundance of the Western Grebe (Aechmophorus occidentalis) in Alberta. M.Sc. thesis. University of Alberta, Edmonton, AB. 82 pp. Erking, S.D. 1994. Feeding Ecology of Fishes. Academic Press, New York, NY.

Page 9-8 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Esler D, Bowman TD, Dean TA, O’Clair CE, Jewett SC, McDonald LL (2000a) Correlates of harlequin duck densities during winter in Prince William Sound, Alaska. Condor 102:920-926 Esler, D. and S.A. Iverson. 2010. Female Harlequin Duck winter survival 11 to 14 years after the Exxon Valdez oil spill. Journal of Wildlife Management. 74(3): 471-478. Esler, D., Ballachy, B., Trust, K.A., Iverson, S.A., Reed, J.A., Miles, A.K., Henderson, J.D., Woodin, B.R., Stegeman, J.J., McAdie, M., Mulcahy, D.M., Wilson, B.W. 2011. Cytochrome P4501, A biomarker indication of timeline of chronic exposure of Barrow’s goldeneyes to residual Exxon Valdez oil. Marine Pollution Bulletin. 62: 609-614. Esler, D., Bowman, T.D., Trust, K.A., Ballachey, B.E., Dean, T.A., Jewett, S.C., O’Clair, C.E. 2002. Harlequin duck population recovery following the Exxon Valdez oil spill: progress, process, and constraints. Marine Ecology Progress Series. 241: 271-286. Esler, D., Trust, K.A., Ballachey, B.E., Iverson, S.A., Lewis, T.L., Rizzolo, D.J., Mulcahy, D.M, Miles, A.K., Woodin, B.R., Stegeman, J.J., Henderson, J.D., Wilson, B.W. 2010. Cytochrome P4501A biomarker induction of oil exposure in harlequin ducks up to 20 years after the Exxon Valdez oil spill. Environmental Toxicology and Chemistry. 29: 1138-1145. Essaid, H.I., Bekins, B.A., Herkelrath, W.N. and Delin, G.N. 2011. Crude Oil at the Bermidji Site: 25 Years of Monitoring, Modeling, and Understanding. Ground Water, 49: 706-726. EVOS Trustee Council 2010. Exxon Valdez Oil Spill Restoration Plan. 2010 Update. Injured Resources and Services. Exxon Valdez Oil Spill Trustee Council, Anchorage, AK. 46pp EVOSTC. 2011. Settlement. Available at: http://www.evostc.state.ak.us/facts/settlement.cfm EVOSTC. 2011. The Reopener. Available at: http://www.evostc.state.ak.us/facts/reopener.cfm Ewins PJ. 1993. Pigeon Guillemot (Cepphus columba), The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Lab of Ornithology; Retrieved from the Birds of North America Online: http://bna.birds.cornell.edu.bnaproxy.birds.cornell.edu/bna/species/049; doi:10.2173/bna.49. Exxon Valdez Oil Spill Trustee Council (EVOSTC). 2010. Exxon Valdez Oil Spill Restoration Plan: 2010 Update Injured Resources and Services. Available at: http://www.evostc.state.ak.us/recovery/status.cfm. Accessed: December 14, 2011. Exxon Valdez Oil Spill Trustee Council. 1994. Exxon Valdez oil spill restoration plan. Exxon Valdez Oil Spill Trustee Council, Anchorage, AK. Available at: www.fakr.noaa.gov/oil/eis/1994RestorationPlan.pdf Exxon Valdez Oil Spill Trustee Council. 2002. Exxon Valdez oil spill restoration plan, Update on Injured Resources and Services. Exxon Valdez Oil Spill Trustee Council, Anchorage, AK. Available at: http://www.evostc.state.ak.us/recovery/status.cfm

July 2012 Page 9-9

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Fall, J.A. 2009. Long-Term Consequences of the Exxon Valdez Oil Spill for Subsistence Uses of Fish and Wildlife. In In Braund, S and J. Kruse (Eds.), Synthesis: Three decades of social science research on socioeconomic effects related to offshore petroleum development in coastal Alaska (MMS OCS Study Number 2009-006, pp. 245-277). Anchorage, AK: Minerals Management Service, Alaska, OCS Region.) Fall, J.A., R. Miraglia, W. Simeone, C. Utermohle, and R.J. Wolfe. 2001. Long-Term Consequences of the Exxon Valdez Oil Spill for Coastal Communities of Southcentral Alaska. James A. Fall (ed.).Technical Report No. 163. OCS Study MMS 2001-032. Final Report for: Sociocultural Consequences of Alaska Outer Continental Shelf Activities: Data Analysis and Integration. Anchorage. Fall, J.A., R.J. Walker, R.T. Stanek, W.E. Simeone, L. Hutchinson-Scarborough, P.Coiley-Kenner, L. Williams, B. Davis, T. Krieg, B. Easley, and D. Koster. 2006. Update of the Status of Subsistence Uses in Exxon Valdez Oil Spill Area Communities, 2003. James A. Fall (ed.). Technical Paper No. 312. Alaska Department of Fish and Game, Division of Subsistence. Juneau. Fall, James A., Rita Miraglia, William Simeone, Charles J. Utermohle, and Robert J. Wolfe. 2001. Long- Term Consequences of the Exxon Valdez Oil Spill for Coastal Communities of Southcentral Alaska. Technical Paper No. 264. Available at: www.subsistence.adfg.state.ak.us/TechPap/tp264.pdf Ford, R.G., M.L. Bonnell, D.H. Varoujean, G.W. Page, B.E. Sharp, and D. Heinemann. 1991. Assessment of direct seabird mortality in Prince William Sound and the Western Gulf of Alaska resulting from the Exxon Valdez oil spill. Unpublished Final Report. Portland, OR: Ecological Consulting, Inc. 221 pp. Fraker, M.A. 2012. Killer whale (Orcinus orca) Deaths in Prince William Sound, Alaska, 1985-1990. Human and Ecological Risk Assessment (in press) Freedman, W., Hutchinson, T.C. 1975. Physical and biological effects of experimental crude oil spills on Low Arctic tundra in the vicinity of Tuktoyatuk, N.W.T., Canada. Canadian Journal of Botany 54:2219-2230. Frost K.J., C.-A. Manen, and T.L. Wade. 1994b. Petroleum hydrocarbons in tissues of harbor seals for Prince William Sound and the Gulf of Alaska. In: T.R. Loughlin (ed), Marine Mammals and the Exxon Valdez. Academic Press, San Diego, CA, USA, pp 331-58 Funk, F., and G. Sandone. 1990. Catch-Age Analysis of Prince William Sound, Alaska, Herring, 1973- 1998. Juneau: Alaska Department of Fish and Games, Fishery Research Bulletin No. 90-01. Galt, J.A., W.J. Lehr, and D.L. Payton. 1991. Fate and transport of the Exxon Valdez oil spill. Environ Sci Technol 25:202-9 Gardmark, A., K. Enberg, J. Ripa, J. Laakso, and V. Kaitala. 2003. The ecology of recovery. Annals Zoologizi Fennici 40:131-144. Garron, J. 2007. End of an era of experimental oil spill sites. Agroborealis. Vol. 39 No. 2: 20 - 27.

Page 9-10 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Garrott RA, Eberhardt LL, Burn DM. 1993. Mortality of sea otters in Prince William Sound following the Exxon Valdez oil spill. Marine Mammal Science 9: 343-359. Garshelis DL, Estes JA. 1997. Sea otter mortality from the Exxon Valdez oil spill: evaluation of an estimate from boat-based surveys. Marine Mammal Science 13: 341-351. Garshelis DL, Johnson CB. 2001. Sea otter population dynamics and the Exxon Valdez oil spill: disentangling the confounding effects. Journal of Applied Ecology 38: 19-35. Geiger, H. J., B. G. Bue, S. Sharr, A. C. Wertheimer, and T. M. Willette. A life history approach to estimating damage to Prince William Sound pink salmon caused by the Exxon Valdez oil spill. Amer. Fish. Soc. Symp., 18: 487–498 (1996). Geraci JR. 1990. Physiologic and toxic effects on cetaceans. In: Geraci JR and St Aubin DJ (eds), Sea Mammals and Oil: Confronting the Risks. Academic Press, San Diego, CA, USA, pp 167-97 Geraci, J.R. and D.J. St. Aubin. 1987. Effects of offshore oil and gas development on marine mammals and turtles. In: D.F. Boesch and N.N. Rabalais (eds), Long-Term Environmental Effects of Offshore Oil and Gas Development. Elsevier Applied Science, New York, NY, USA, pp 587-617 Gill, Duane A. and Liesel A. Ritchie. 2006. “Community Responses to Oil Spills: Lessons to be Learned from Technological Disaster Research” in Brewer, Reid, Editor. The Selendang Ayu Oil Spill: Lessons Learned. Conference Proceedings, Augusut 16-19, 2005 – Unalaska, Alaska, Available at: www.iccopr.uscg.gov as 2006_Selendang_Ayu_Oil_Spill_Proceeding.pdf Gillfillan, ES, D. Page, E. Harner, and P.Boehm. 1995. Shoreline ecology program for Prince William Sound Alaska following the EVOS - Biology. In: EVOS: Fate and Effects in Alaskan Waters. ASTM STP 1219 Philad. Pa. Gilpin, M.E. and M.E. Soulé. 1986. Minimum viable populations: processes of species extinction. In: M.E. Soulé (ed), Conservation Biology: The Science of Scarcity and Diversity. Sinauer Associates, Sunderland, MA, USA, pp 18-35 Glemarec, M. and Hussenot, E. (1981) Definition d’une succession ecologique en milieu meuble anormalement enrichi en matieres organiques a la suite de la catastrophe de l’Amoco Cadiz. In Amoco Cadiz : Consequences d'une Pollution Accidentelle par les Hydro-carbures; Fates and Effects of the oil spill, pp. 499-525. Centre National pour l’Exploitation des Oceans, Paris. Goldberg, H. 2006. Pine River: 2005 Assessment- Residual oil survey and snorkel survey. Prepared by Arc Environmental Ltd. Kamloops, BC. Goldberg, Harry. "Pine River 2011 Fisheries Update: Status of Recovery Post-2000 Pipeline Rupture." Enbridge, 2011. Goldstein, R.M., and T.P. Simon. 1999. Toward a united definition of guild structure for feeding ecology of North American freshwater fishes. Pages 123 – 202 in T.P. Simon, editor. Assessing the sustainability and biological integrity of water resources using fish communities. CRC Press, New York, New York, USA.

July 2012 Page 9-11

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Golet, G.H., Kuletz, K.J., Roby, D.D., Irons, and D.B. 2000. Adult prey choice affects chick growth and reproductive success of pigeon guillemots. Auk 117: 82-91. Golet, G.H., P.E. Seiser, A.D. McGuire, D.D. Roby, J.B. Fischer, K.J. Kuletz, D.B. Irons, T.A. Dean, S.C. Jewett, and S.H. Newman. 2002. Long-term direct and indirect effects of the ‘Exxon Valdez’ oil spill on pigeon guillemots in Prince William Sound, Alaska. Marine Ecology Progress Series. 241: 287-304. Gurska, J., Wang, W., Gerhardt-Hogenkamp, K.E., Khalid, A., Isherwood, D., Huang, X.D., Glick, B.R., Greenberg, B.M. 2009. Successful field test of a multi-process phytoremediation system at a petroleum sludge contaminated land farm. Environmental Science and Technology. 43: 4472- 4479. Haggarty, J., Wooley, C., Erlandson, J., Crowell, A., 1991. The 1990 Exxon Cultural Resource Program: Site Protection and Maritime Cultural Ecology in Prince William Sound and the Gulf of Alaska. Exxon Company, Anchorage, USA. Hampton, Steve, Angie Montalvo, Damian Higgins, and Pat Sollberger. "Assessment of Natural Resource Damages as a Result of the East Walker River Oil Spill on December 30 2000." edited by California Department of Fish and Game - Office of Spill Prevention and Response, U.S. Fish and Wildlife Service and Nevada Division of Wildlife, 2002. Hanna, S.R. and P.J. Drivas. 1993. Modeling VOC emissions and air concentrations from the Exxon Valdez oil spill. Air Waste 43:298-309 Hargreaves, N. B., and R. J. Lebrasseur. 1985. Species selective predation on juvenile pink (Oncorhynchus gorbuscha) and chum salmon (O. keta) by coho salmon (O. kisutch). Canadian Journal of Fisheries and Aquatic Sciences 42:659–668. Harper, K.A., Kershaw G.P. 1996. Natural Revegetation on Borrow Pits and Vehicle Tracks in Shrub Tundra, 48 Years Following Construction of the CANOL No.1 Pipeline, N.W.T. Canada. Arctic and Alpine Research 28(2): 163-171. Hartung, R. 1967. Energy metabolism in oil-covered ducks. Journal of Wildlife Management 31:798-804. Harvey, J.T. and M.E. Dahlheim. 1994. Cetaceans in oil. In: T.R. Loughlin (ed), Marine Mammals and the Exxon Valdez. Academic Press, San Diego, CA, USA, pp 257-64 Harvey, S., P.J. Sharp, and J.G. Phillips. 1982. Influence of ingested petroleum on the reproductive performance and pituitary-gonadal axis of domestic ducks (Anas platyrhynchols). Comparative Biochemistry and Physiology 72:83-89. Harwell, MA, and JH Gentile. 2006. Ecological significance of residual exposure and effects from the Exxon Valdez oil spill. Integrated Environmental Assessment and Management 2:204-246. Hatler, D. F., D. W. Nagorsen, and A. M. Beal. 2008. Carnivores of British Columbia. Royal BC Museum Handbook, UBC Press, Vancouver. 416 pages.

Page 9-12 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Hawkins, S.J. and Southward, A.J. (1992) The Torrey canyon oil spill: recovery of rocky shore communities. In: Chapter 13 of Restoring the Nation’s Marine Environment (Thayer, G.W., ed.). pp.583-619. National Oceanic and Atmospheric Administration, Washington D.C. Hay, D.E. 1985. Reproductive Biology of Pacific Herring (Clupea harengus pallasi). Canadian Journal of Fisheries and Aquatic Sciences 42 (Suppl. 1): 111-26. Hay, D.E. and P.B. McCarter. 2011. Herring spawning areas of British Columbia: A review, geographical analysis and classification. Revised and updated MS Rept. 2019: Available online: http://www.pac.dfo-mpo.gc.ca/science/species-especes/pelagic-pelagique/herring- hareng/herspawn/pages/acknowl1-eng.htm Hay, D.E., McCarter, P.B., and Grant, D. 1995. Nestucca oil spill and impacts on herring eggs. In Proceedings of the Seventh Pacific Coast Herring Workshop, January 27–28, 1994. Edited by D.E. Hay and P.B. McCarter. Can. Tech. Rep. Fish. Aquat. Sci. No. 2060. pp. 95–100. Heard, W. H. Life history of pink salmon, pp. 119–230 In: Pacific Salmon Life Histories (Groot, C., and L. Margolis, Eds.). Vancouver: University of British Columbia Press (1991). Heinemann, D. 1993. How long to recovery for murre populations, and will some colonies fail to make the comeback? In: Program and abstract book, Exxon Valdez Oil Spill Symposium, February 2-5, 1993, Anchorage, AK. Anchorage: Oil Spill Trustee Council: 139-141. Heintz, R.A., J.W. Short, and S.D. Rice. 1999. Sensitivity of fish embryos to weathered crude oil: Part II. Increased mortality of pink salmon (Oncorhynchus gorbuscha) embryos incubating downstream from weathered Exxon Valdez crude oil. Environ. Toxicol. Chem. 494-503. Hershberger P, Hart A, Gregg J et al (2006) Dynamics of viral hemorrhagic septicemia, viral erythrocytic necrosis, and ichthyophoniasis in confined juvenile Pacific herring Clupea pallasi. Dis Aquat Org 70:201-208 Hershberger, P.K., N.E. Elder, J. Wittouck, K. Stick and R.M. Kocan. 2005. Abnormalities in Larvae from the Once-Largest Pacific Herring Population in Washington State Result Primarily from Factors Independent of Spawning Location. Transactions of the American Fisheries Society 134: 326-337. Higgins, Damian K. . "Assessment of Damages to Natural Resources in the East Walker River from the Advanced Fuel Filtration Spill: Water, Sediment, and Fish Tissue Analysis." edited by Interior, 26. Reno, NV: U.S. Fish and Wildlife Services Nevada Fish and Wildlife Office 2002. Hilderbrand, G.V., C.C. Schwartz, C.T. Robbins, M.E. Jacoby, T.A. Hanley, S.M. Arthur, and C. Servheen. 1999a. The importance of meat, particularly salmon, to body size, population productivity, and conservation of North American brown bears. Canadian Journal of Zoology 77: 132-138. Hilderbrand, G.V., S.G. Jenkins, C.C. Schwatz, T.A. Hanley, and C.T. Robbins, C.T. 1999b. Effect of seasonal differences in dietary meat intake on changes in body mass and composition in wild and captive brown bears. Canadian Journal of Zoology 77: 1623-1630.

July 2012 Page 9-13

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Hodison, P.V., I. Ibrahim, S. Zambon, A. Ewert, and K. Lee. 2002. Bioavailability of fish to sediment PAH as an indicator of the success of in situ remediation treatments at an experimental oil spill. Bioremediation Journal. 6(3):297 – 313. Holland, M., and E. Bluemink. 2010. Pump station oil spill may close Alaska pipeline until Friday. Anchorage Daily News. Available: http://www.adn.com/2010/05/26/1295235/responders-move- slowly-on-alaska.html. Accessed 20 June 2012. Holt, Sune. 1987. The Effects of Crude and Diesel Oil Spills on Plant Communities at Mesters Vig. Northeast Greenland. Arctic and Alpine Research. Restoration and Vegetation Succession in Circumpolar Lands: Seventh Conference of the Comité Arctique International. 19(4): 490-497. Hom, T., U. Varanasi, J.E. Stein, et al. 1996. Assessment of the exposure of subsistence fish to aromatic compounds after the Exxon Valdez oil spill. Am Fish Soc Symp 18:856-66 Hoover-Miller, A.., K.R. Parker and J.J. Burns. 2001. A reassessment of the impact of the Alaska Exxon Valdex oil spill on harbor seals (Phoca vitulina richardsi) in Prince William Sound. Marine Mammal Science, 17:111-135. Horne, A.J., and C.R. Goldman. 1994. Limnology. 2nd Edition. McGraw-Hill, Inc. Hose, J.E., M.D. McGurk, G.D. Marty, D.E. Hinton, E.D. Brown, and T. Baker. 1996. Sublethal Effects of the Exxon Valdez Oil Spill on Herring Embryos and Larvae: Morphological, Cytogenetic, and Histopathological Assessments, 1989. Canadian Journal of Fisheries and Aquatic Sciences 53: 2355-65. Houghton, J.P., Gilmour, R.H., Lees, D.C., Driskell, W.B., Lindstrom, S.C., Mearns, A., 1997. Prince William Sound intertidal biota seven years later: Has it recovered? Proceedings of 1997 International Oil Spill Conference, 679–686 Huang, X.D., El-Alawi, Y., Gurska, J., Glick, B.R. 2005. A multi-process phytoremediation system for decontamination of persistent total petroleum hydrocarbons (TPHs) from soils. Microchemical Journal 81: 139-147. Hugget, RJ, JM Neff, JJ Stegeman, B Woodin, KR Parker, and JS Brwon. 2006. Biomarkers of PAH exposure in an intertidal fish species from Prince William Sound, Alaska: 2004-2005. Environmental Science and Tecnology 40:6513-6517. Hulson, P.-J.F., S.E. Miller, T.J. Quinn, G.D. Marty, S.D. Moffitt, and F. Funk. 2008. Data Conflicts in Fishery Models: Incorporating Hydroacoustic Data into the Prince William Sound Pacific Herring Assessment Model. ICES Journal of Marine Science: Journal du Conseil 65: 25-43. Hutchinson, T.C. 1984. Recovery or Arctic and sub-arctic vegetation nine summers after crude diesel spills. Environmental Studies No. 22. Indian and Northern Affairs Canada, Ottawa. Hutchinson, T.C. and W. Freedman. 1978. Effects of experimental crude oil spills on subarctic boreal forest vegetation near Norman Wells, NWT, Canada. Canadian Journal of Botany 56:2424-2433.

Page 9-14 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Impact Assessment Inc. 2001. Exxon Valdez Oil Spill, Clean-up, and Litigation: A Collection of Social- Impacts Information and analysis, Final Report, Volume I: final Comprehensive Report. Available at: www.gomr.boemre.gov/Pl/PDFImages/ESPIS/3/3403.pdf Impact Assessment Inc. 2011. Critical Human Dimensions of Maritime Oil Spills As Identified through Examination of the Selendang Ayu Incident. OCS Study BOEM 053-2011. Available at: www.ntis.gov/search/product.aspx?ABBR+PB2012101976 Impact Assessment, Inc. 2010 Social and Economic Assessment of Major Oil Spill Litigation Settlement – Final Baseline Report – Prepared for U.S. Department of the Interior Bureau of Ocean Energy, management, Regulation and Enforcement. Contract Number 0103CT72619. Available at: http://www.alaska.boemre.gov/reports/2010rpts/2010rpts.HTM as Integral Consulting Inc. 2006. Information Synthesis and Recovery Recommendations for Resources and Services Injured by the Exxon Valdez Oil Spill. Exxon Valdez Oil Spill Restoration Project Final Report [Restoration Project 060783]. Exxon Valdez Oil Spill Trustee Council, Anchorage, AK. International Bank for Reconstruction and Development. 1995. Report No. T-6578-RU, Emergency Oil Spill Recovery and Mitigation Project - Technical Annex, April, 1995 International Bank for Reconstruction and Development. 1995b. “Miracle in the Arctic: Oil Spill Clean- up is Successful”, Press Release No:96/009/ECA, October 1995. Available at: http://go.worldbank.org/O3GZHG5HJ1 IPIECA (1999) Biological Impacts of Oil Pollution: Sedimentary Shores. London: International Petroleum Industry Environmental Conservation Association. IPIECA Report Series, Vol. 9., 20pp. Irons, D.B., S.J. Kendall, W.P. Erickson, L.L. McDonald, and B.K. Lance. 2000. Nine years after the “Exxon Valdez” oil spill: effects on Marine Bird Populations in the Prince William Sound, Alaska. The Condor. 102(4): 723-737. ITOPF. 2011. ITOPF’s Spill Information Pack. Available at: http://www.itopf.com/information-services/data-and-statistics/statistics/ ITOPF. 2011b. Case Histories: the Exxon Valdez. Available at http://www.itopf.com/information- services/data-and-statistics/case-histories/elist.html Iverson, S.A. and D. Esler. 2010. Harlequin duck population injury and recovery dynamics following the 1989 Exxon Valdez oil spill. Ecological Applications. 20: 1993-2006. Jackson, ST, and RJ Hobbs. 2009. Ecological restoration in light of ecological history. Science 325:567- 569. Jenkins, T.F., L.A. Johnson, C.M. Collins, and T.T. McFadden. 1978. The physical, chemical and biological effects of crude spills on black spruce forest, interior Alaska. Arctic 31:305-323. Jenssen, B.M., and M. Ekker. 1991. Effects of plumage contamination with crude oil dispersant mixtures on thermoregulation in common eiders and mallards. Archives of Environmental Contamination and Toxicology 20:398-403.

July 2012 Page 9-15

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Jewett SC, Dean TA, Smith RO, Stekoll M, Haldorson LJ, Laur DR, McDonald L. 1995. The effects of the Exxon Valdez oil spill on shallow subtidal communities in Prince William Sound, Alaska, 1989–93. Anchorage (AK): Alaska Department of Fish and Game, Habitat and Restoration Division. Restoration Project 93047 (Subtidal Study 2A) Final Report. 339 p. Jewett SC, Dean TA, Woodin BR, Hoberg MK, Stegeman JJ. 2002. Exposure to hydrocarbons 10 years after the Exxon Valdez oil spill: Evidence from cytochrome p4501A expression and biliary FACs in nearshore demersal fishes. Mar Environ Res 54:21–48. Jewett SC, Dean TA. 1997. The effects of the Exxon Valdez oil spill on eelgrass communities in Prince William Sound, Alaska 1990–95. Anchorage (AK): Alaska Department of Fish and Game, Habitat and Restoration Division. Restoration Project 95106 Final Report. 291 p. Jewett, S.C., Dean, T.A., Smith, R.O. and Blanchard, A. 1999. Exxon Valdez oil spill: impacts and recovery in the soft-bottom benthic community in and adjacent to eelgrass beds. Marine Ecology Progress Series, 185: 59–83. Johansson, S., U. Larsson, and P. Boehm. 1980. The Tsesis oil spill impact on the pelagic ecosystem. Mar. Pollut. Bull. 11:284-293. Johnson, S.W., M.G. Carls, R.P. Stone, C.C. Brodersen, and S.D. Rice. 1997. Reproductive Success of Pacific Herring, Clupea pallasi, in Prince William Sound, Alaska, Six Years after the Exxon Valdez Oil Spill. Fishery Bulletin 94: 748-61. Jones, HP, and OJ Schmitz. 2009. Rapid recovery of damaged ecosystems. PLoS ONE 4:e5653. Doi:10.1371/journal.pone.0005653. Kamath, R., Rentz, J.A., Schnoor, J.L., Alvarez, P.J.J. 2004. Phytoremediation of hydrocarbon- contaminated soils: principles and applications. In: Vazquez-Duhalt, R., Quintero-Ramirez, R. (Eds). Petroleum Biotechnology: Developments and Perspectives Studies in Surface Science and Catalysis. Elsevier Science, Oxford, UK. 447-478. Kapustka LA, and WG Landis. 1998. Ecology: The science versus the myth. Human and Ecological Risk Assessment 4:829-838. Kelly, Jeffrey R. 2002. Exxon Valdez Oil Spill, 10-Day Clean-up Timeline. Available at: http://library.thinkquest.org/10867/clean-up/timeline.shtml Kelty, Frank. 2006. “Impacts from Oil Spills on Fisheries and to the Seafood Industry That Supports the City of Unalaska” in Brewer, Reid, Editor. The Selendang Ayu Oil Spill: Lessons Learned. Conference Proceedings, August 16-19, 2005 – Unalaska, Alaska, Available at: www.iccopr.uscg.gov as 2006_Selendang_Ayu_Oil_Spill_Procedding.pdf Kennedy CJ, Farrell AP, Sanders S (1999) Survival, performance and reproduction in the pacific herring, clupea harems pallasi: effects of environmental contamination, viral hemorrhagic septicemia virus and ichthyophonus hofer. In: Kocan RM, Marty GD, Kennedy CJ (eds) Investigations of disease factors affecting declines of Pacific herring populations in Prince William Sound, Final Report, Restoration Project 97162, Exxon Valdez Oil Spill Trustee Council, Anchorage, pp III-1- III-40, http://www.evostc.state.ak.us/Store/FinalReports/1998-98162-Final.pdf

Page 9-16 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Kennedy, J., J.E. Skjaeraasen, R.D.M. Nash, A. Thorsen, A. Slotte, T. Hansen, and O.S. Kjesbu. 2010. Do capital breeders like Atlantic herring (Clupea harengus) exhibit sensitive periods of nutritional control on ovary development and fecundity regulation? Can. J. Fish. Aquat. Sci. 67:16-27. Kingston, P.F., Dixon, I.M.T., Hamilton, S., Moore, C.G. and Moore, D.C. (1997) Studies on the response of intertidal and subtidal marine benthic communities to the Braer oil spill. In: The Impact of an Oil Spill in Turbulent Waters: The Braer. The Stationery Office Limited, Edinburgh, pp. 209-233 Kingston, PF. 2002. Long-term environmental impact of oil spills. Spill Science and Technology Bulletin 7:53-61. Knapp, Gunnar. 2005. Oil Spills and Fisheries Markets: Potential Effects and Mitigation Strategies. Presentation to the Aleutian Life Forum, Effects of Oil on Fisheries, Unalaska, Alaska. Available at: www.iser.uaa.alaska.edu, as Knapp_Oil_Spill_Market_Effects.pdf. Kocan RM, Hershberger PK, Mehl P et al (1999a) Pathogenicity of Ichthyophonus hoferi for laboratory- reared Pacific herring Clupea pallasi and its early appearance in wild Puget Sound herring. Dis Aquat Organ 35:23-29 Kocan RM, Hershberger PK, Winton JR, et al (1999b) Controlled field and laboratory studies on VHS and Ichthyophonus in Pacific herring. In: Kocan RM, Marty GD, Kennedy CJ (eds) Invesigations of disease factors affecting declines of Pacific herring populations in Prince William Sound, Final Report, Restoration Project 97162, Exxon Valdez Oil Spill Trustee Council, Anchorage, pp II-i- II-73, http://www.evostc.state.ak.us/Store/FinalReports/1998-98162-Final.pdf Kohout, Jneifer and Norman Meade. 2008. Preassessment Data Report #12: Potential Human Use Losses Associated with the Selendang Ayu Oil Spill. Available at: http://www.docstoc.com/docs/31096047/Subsistence Krebs, C. T., Burns, K. A., 1977. Long-term effects of an oil spill on populations of the salt-marsh crab Uca pugnax. Science 197, 484-487. Krebs, CT and Burns, KA (1977). Long-term effects of an oil spill on the salt marsh crab Uca pugnax. Science, 197, 484-487. Kubach, Kevin M., Mark C. Scott, and James S. Bulak. "Recovery of a Temperate Riverine Fish Assemblage from a Major Diesel Oil Spill." Freshwater Biology 56, no. 3 (2011): 503-18. Lance BK, Irons DB, Kendall SJ, McDonald LL. 2001. An evaluation of marine bird population trends following the Exxon Valdez oil spill, Prince William Sound, Alaska. Marine Pollution Bulletin 42: 298-309. Landis, WG. 2007. The Exxon Valdez oil spill revisited and the dangers of normative science. Integrated Environmental Assessment and Management 3:439-441. Leatherwood, S., C.O. Matkin, J.D. Hall, et al. 1990. Killer whales, Orcinus orca, photo-identified in Prince William Sound, Alaska, 1976-1987. Can Field-Natur 104:362-71

July 2012 Page 9-17

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Lee, K. et al. (1999) Natural recovery reduces impact of the 1970 Arrow oil spill. In: Proceedings of the 1999 International Oil Spill Conference. 5pp. Lee, M.D., and C.H. Ward. 1985. Biological methods for the restoration of contaminated aquifers. Environmental Toxicology and Chemistry 4:743-750. Lee, RF and Page, DS (1997) Petroleum hydrocarbons and their effects in subtidal regions after major oil spills. Marine Pollution Bulletin, 34 (11), 928-940. Lees DC and Driskell WB (2007) Assessment of Bivalve Recovery on Treated Mixed-Soft Beaches in Prince William Sound. Exxon Valdez Oil Spill Restoration Project Final Report. NOAA, Alaska. 111pp. Lees, D. (2011) EVOS Project 10100574 - Re-assessment of Bivalve Recovery on Treated Mixed-Soft Beaches in Prince William Sound. Annual Project Report to Exxon Valdez Oil Spill Trustee Council. 5pp. Lewis, J.P., and D.G. Calkins. 1995. Assessment of the Exxon Valdez oil spill on the Sitka black-tailed deer in Prince William Sound and the Kodiak Archipelago. Exxon Valdez Oil Spill State/Federal Natural Resource Damage Assessment Annual Report, Alaska Department of Fish and Game, Anchorage, AK. Lewis, S.J., and R.A. Malecki. 1984. Effects of egg oiling on larid productivity and population dynamics. Auk 101:584-592. Linden, Olof, Ragnar Elmgren, and Paul Boehm. "The Tsesis Oil Spill: Its Impact on the Coastal Ecosystem of the Baltic Sea." Ambio 8, no. 6 (1979): 246-53. Lindstrom, Jon E., Ronald P. Barry, and Joan F. Braddock. "Long-Term Ffects on Microbial Communities after a Subarctic Oil Spill." Soil Biology and Biochemistry 31 (1999): 1677-89. Lipscomb TP, Harris RK, Rebar AH, Ballachey BE, Haebler RJ. 1994. Pathology of sea otters. In Loughlin TR (ed.). Marine Mammals and the Exxon Valdez. Academic Press, San Diego. pp 265- 280. Little DI, Rhind P, Jones R, Bennett I and Moore J (1997) CCW's shoreline oil distribution surveys following the Sea Empress spill. In: Proceedings of the 1997 International Oil Spill Conference. Litzow MA, Piatt JF, Prichard AK, Roby DD. 2002. Response of pigeon guillemots to variable abundance of high-lipid and low-lipid prey. Oecologia 132: 286-295. Los Angeles Times. 1992. Alyeska Settles Valdez Spill Suit for $32 Million. Available at: http://articles.latimes.com/1992-11-26/busines/fi-1414_1_exxon-valdez-oil-spill Loughlin, T.R. 1994. Tissue hydrocarbon levels and the number of cetaceans found dead after the spill. In: T.R. Loughlin (ed), Marine Mammals and the Exxon Valdez. Academic Press, San Diego, CA, USA, pp 359-70 Lowry, L.F., K.J. Frost, and K.W. Pitcher. 1994. Observations of oiling of harbor seals in Prince William Sound. In: T.R. Loughlin (ed), Marine Mammals and the Exxon Valdez. Academic Press, San Diego, CA, USA, pp 209-25

Page 9-18 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Lytle, David A., and Barbara L. Peckarsky. "Spatial and Temporal Impacts of a Diesel Fuel Spill on Stream Invertebrates." Freshwater Biology 46 (2001): 693-704. Maccarone, A.D., and J.N. Brzorad. 2000. Wading bird foraging: response and recovery from an oil spill. Waterbirds 23:246-257. MacDonald, BA and Thomas, MLH (1982) Growth reduction in the soft-shell clam Mya arenaria from a heavily oiled lagoon in Chedabucto Bay, Nova Scotia. Marine Environmental Research, 6, 145- 156. MacDonald, D.D., C.G. Ingersoll, and T.A. Berger. 2000. Development and evaluation of consensus- based sediment quality guidelines for freshwater ecosystems. Arch. Environ. Contam. Toxicol. 39:20-31. Mackie P.R., Hardy R., Butler E.I., Holligan P.M., Spooner M.F. (1978) Early samples of oil in water and some analyses of zooplankton. Marine Pollution Bulletin, Volume 9(11), 296–297. Macko, S.A., and S.M. King. 1980. Weathered oil: effect on hatchability of heron and gull eggs. Bulletin of Environmental Contaminants and Toxicology 25:316-320. Maki, A. W., E. L. Brannon, L. G. Gilbertson, L. L. Moulton, and J. R. Skalski. An assessment of oil spill effects on pink salmon populations following the Exxon Valdez oil spill. Part 2: Adults and escapement, pp. 585–625. In: Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters (Wells, P. G., J. N. Butler, and J. S. Hughes, Eds.) Philadelphia, PA: American Society for Testing and Materials (1995). Marty, G. D., R. A. Heintz, and D. E. Hinton. Histology and teratology of pink salmon larvae near the time emergence from the gravel substrate in the laboratory. Can. J. of Zoology, 75: 978–988 (1997a). Marty, G.D. 2007. Evidence That the Exxon Valdez Oil Spill Did Not Cause the 1993 Disease Epidemic in the Pacific Herring Population of Prince William Sound Alaska. In Prince William Sound Herring: An Updated Synthesis of Population Declines and Lack of Recovery. Edited by S.D. Rice and M.G. Carls, Anchorage: Exxon Valdez Oil Spill Trustee Council. Marty, G.D., E.F. Freiberg, T.R. Meyers, J. Wilcock, T.B. Farver, and D.E. Hinton. 1998. Viral Hemorrhagic Septicemia Virus, Ichthyophonus hoferi, and Other Causes of Morbidity in Pacific Herring Clupea pallasi Spawning in Prince William Sound, Alaska, USA. Diseases of Aquatic Organisms 32: 15-40. Matkin, C.O., E.L. Saulitis, G.M. Ellis et al. 2008. Ongoing population-level impacts on killer whales Orcinus orca following the ‘Exxon Valdez’ oil spill in Prince William sound, Alaska. Mar Ecol Progr Ser 256:269-81 Matkin, C.O., G.M. Ellis, M.E. Dahlheim, et al. 1994. Status of killer whales in Prince William Sound, 1985-1992. In: T.R. Loughlin (ed), Marine Mammals and the Exxon Valdez. Academic Press, San Diego, CA, USA, pp 141-62

July 2012 Page 9-19

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Mauseth, G., Svarny-Livingston, S. Arnold, G. Erickson. 2008. The M/V Selendang Ayu Oil Spill Subsistence Seafood Sampling Program and Public Health Evaluation Process. Proceedings of the 2008 International Oil Spill Conference, May 2008, Savannah, GA. McDowell Group. 1990. An Assessment of the Impact of the Exxon Valdez Oil Spill on the Alaska Tourism Industry. Phase I, Initial Assessment. Available at: http://www.evostc.state.ak.us/facts/economic.cfm McGurk, M.D., and E. D. Brown. 1996. Egg-Larval Mortality of Pacific Herring in Prince William Sound, Alaska, after the Exxon Valdez Oil Spill. Canadian Journal of Fisheries and Aquatic Sciences 53: 2343-54. McKendrick, J.D. and W.M.W. Mitchell. 1978. Fertilizing and seeding oil-damaged arctic tundra to effect vegetation recovery, Prudhoe Bay, Alaska. Arctic 31(3):296-304. McKendrick, J.D. and WM. W. Mitchell 1978. Fertilizing and seeding oil-damaged Arctic tundra to effect vegetation recovery Prudhoe Bay, Alaska. Arctic Vo. 31 No. 3 296 – 304. Meador, J.P. 2003. Bioaccumulation of PAHs in marine invertebrates. In Douben, P.E.T., ed., PAHS: An Ecological Perspective, Ecological and Environmental Toxicology Series. John Wiley and Sons, Ltd., West Sussex, England.147-171 pp. Merkl, N. Schultze-Kraft, R, Infante, C. 2005. Assessment of tropical grasses and legumes for phytoremediation of petroleum-contaminated soils. Water, Air, and Soil Pollution 165:195-209. Michael, A.D. 1977. The effects of petroleum hydrocarbons on marine populations and communities, pp. 129-137. In D.A. Wolfe, ed. Fate and Effects of Petroleum Hydrocarbons in Marine Organisms and Ecosystems. Pergamon, New York Michael, A.D., Van Raalte, C.R. and Brown, L.S. (1975) Long-term effects of an oil spill at West Falmouth, Massachusetts. In: 1975 Conference on Prevention and Control of Oil Pollution: Proceedings. Washington, D.C.: American Petroleum Institute. Michel J, Nixon Z, Hayes MO, Short J, Irvine G, Betenbaugh D, Boring C and Mann D. 2010. Distribution of Subsurface Oil from the Exxon Valdez Oil Spill, Exxon Valdez Oil Spill Restoration Project Final Report [Restoration Project 070801]. National Oceanic and Atmospheric Administration, Juneau, AK Michigan Department of Environmental Quality (MDEQ). 1990. SWAS Procedure 51 – Qualitative Biological and Habitat Survey Protocols for Wadeable Streams and Rivers, April 24, 1990. Revised June 1991, August 1996, January 1997, May 2002, and December 2008. Michigan Department of Environmental Quality (MDEQ). 2012. A biological survey of sites on the Kalamazoo River and Talmadge Creek near the Enbridge oil spill in Marshall Calhoun County, Michigan August 2011. Water Resources Division February 2012 Staff Report. MI/DEQ/WRD- 12/012. Mills, Michael J. 1992. Alaska Sport fishing in the Aftermath of the Exxon Valdez Oil Spill. Alaska Department of Fish and Game. Special Publication No. 92-5. Available at: http://www.evostc.state.ak.us/facts/economic.cfm

Page 9-20 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Moles, A., and S. D. Rice. Effects of crude oil and naphthalene on growth, caloric content and fat content of pink salmon juveniles in sea water. Trans. Am. Fish. Soc., 112: 205–211 (1983). Monson DH, Doak DF, Ballachey BE, Bodkin JL. 2011. Could residual oil from the Exxon Valdez spill create a long-term population ‘sink’ for sea otters in Alaska? Ecological Applications 21: 2917- 2932. Monson DH, Doak DF, Ballachey BE, Johnson A, Bodkin JL. 2000. Long-term impacts of the Exxon Valdez oil spill on sea otters, assessed through age-dependent mortality patterns. Proceedings of the National Academy of Sciences 97: 6562-6567. Moore, C.G., Harries, D.B., Ware, F.J. (1997) The impact of the Sea Empress oil spill on the sandy shore meiofauna of south west Wales. A report to the Countryside Council for Wales from Heriot-Watt University. 79pp. Moore, J. (2006a) Long term ecological impacts of marine oil spills. In: Proceedings of the Interspill 2006 conference, held at London ExCeL, 21-23 March 2006. Moore, J. (2006b) State of the marine environment in south west Wales 10 years after the Sea Empress oil spill. A report to the Countryside Council for Wales from Coastal Assessment, Liaison and Monitoring, Cosheston, Pembrokeshire. 24pp plus vi. Moore, J.J., Taylor, P.M.H. and Hiscock, K. (1995). Rocky shores monitoring programme. In: Monitoring at an Oil Terminal: The Shetland Experience (eds. Dunnet, G.M. and McIntyre, A.D.). Proceedings of the Royal Society of Edinburgh, 103B, 181-200. Morin, P.A., F.I. Archer, A.D. Foote et al. 2010. Complete mitochondrial genome phylogenetic analysis of killer whale (Orcinus orca) indicates multiple species. Genome Research 20:908-916 Mosbech, A. (ed.) 2002. Potential Environmental impacts of oil spills in Greenland. An assessment of information status and research needs. National Environmental Research Institute, Denmark. 118 pp. – NERI Technical Report No. 415. Moulton, L. L. Effects of oil-contaminated sediments on early life stage and egg viability of pink salmon in Prince William Sound Alaska, pp. 147–155. In: Proceedings of the 17th Northeast Pacific Pink and Chum Salmon Workshop 1995, (Fuss, H. and G. Graves, Eds.). Northwest Fisheries Commission and Washington Department of Fish and Wildlife, Olympia, WA (1996). Murphy, S.M., and T.J. Mabee. 2000. Status of black oystercatchers in Prince William Sound, Alaska, nine years after the Exxon Valdez oil spill. Waterbirds: The International Journal of Waterbirds. 23: 204-213. Murphy, S.M., R.H. Day, J.A. Wiens and K.R. Parker. 1997. Effects of the Exxon Valdez oil spill on birds: Comparisons of pre- and post-spill surveys in Prince William Sound, Alaska. Condor. 99: 299-313. National Academy of Sciences 1985. Oil in the Sea: Inputs, Fates and Effects. National Academy Press, Washington D.C. 280pp.

July 2012 Page 9-21

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

National Academy of Sciences 2003. Oil in the sea III: Inputs, fates and effects. Washington: National Academies Press. National Transportation Safety Board 2012. Emergency and Environmental Response Group Chairman’s Factual Report. Available at: http://dms.ntsb.gov/pubdms/search/document.cfm?docID=362447&docketID=49814&mkey=767 66 Nedunuri, K.V., Govindaraju, R.S., Banks, M.K., Schwab, A.P., Chen, Z. 2010.Evaluation of phytoremediation for Field-Scale Degradation of Total Petroleum hydrocarbons. Journal of Environmental Engineering. 126(6): 483-490. Nedunuri, K.V., Lowell, C., Meade, W., Vonderheide, A.P., J.R. Shann. 2010. Management practices and phytoremediation by native grasses. International Journal of Phytoremediation, 12: 200-214. Neff JM, Burns WA (1996) Estimation of polycyclic aromatic hydrocarbon concentrations in the water column based on tissue residues in mussels and salmon: an equilibrium partitioning approach. Environ Toxicol Chem 15(12):2240-2253 Nellbring, S, S Hansson, G Aneer, and L Westin. 1980. Impact of oil on local fish fauna. P 193-201. In: JJ Kinemann, R Emgren, and S Hansson (eds) The Tsesis Oil Spill. U.s. Department of Commerce, NOAA, Wahington, DC Nikitik C.C.S. and Robinson A.W. (2003), Patterns in benthic populations in the Milford Haven waterway following the Sea Empress oil spill with special reference to amphipods. Mar. Pollut. Bull. 46, 1125-1141. NOAA/BML. 2008. The 2007 : Assessing toxic injury to Pacific herring embryos and larvae in the San Francisco Estuary. Northwest Fisheries Science Center, National Marine Fisheries Service, and National Ocean and Atmosphere Administrations: Environmental Conservation Division; and Ecotoxicology and Environmental Science and Policy and Aquatic Resources Group, Bodega Marine Laboratory. Available from http://rosebud.journalism.berkeley.edu/~j200/mission/2009/10/20091022_herring/site/reference/c osco_busan_oil_spill.pdf. Norcross, B.L., and E.D. Brown. 2001. Estimation of First-Year Survival of Pacific Herring from a Review of Recent Stage-Specific Studies. In Proceedings of the 18th Lowell Wakefield Fisheries Symposium, Herring 2000: Expectations for a New Millennium, Edited by F. Funk, J. Blackburn, D. Hay, A. J. Paul, R. Stephenson, R. Toresen and D. Witherell, 535-558. Fairbanks: University of Alaska Sea Grant. Norcross, B.L., E.D. Brown, R.J. Foy, M. Frandsen, S.M. Gay, T.C. Kline, D.M. Mason, E.V. Patrick, A. J. Paul, and K.D.E. Stokesbury. 2001. A Synthesis of the Life History and Ecology of Juvenile Pacific Herring in Prince William Sound, Alaska. Fisheries Oceanography 10 (2001): 42-57. NRC (National Research Council) 2003. Oil in the Sea III: Inputs, Fates, and Effects. Ocean Studies Board and Marine Board, Divisions of Earth and Life Science Studies and Transportation Research Board, National Research Council of the National Academies. Washington, D.C. 278 pp.

Page 9-22 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

NRC (National Research Council). 2005. Oil Spill Dispersants, Efficacy and Effects. National Academies Press, Washington, DC, USA O’Donnel, C., and J. Fjeldsa. 1997. Grebes-status study and conservation action plan. IUCN/SSC Grebe Specialist Group. IUCN. Gland, Switzerland and Cambridge, UK. Oritsland, N.A., F.R. Engelhardt, F.A. Juck, R.J. Jurst, and P.D. Watts. 1981. Effects of crude oil on polar bears. Northern Affairs Program, Environmental Studies No. 24. Ottawa, Ontario, Canada. Owens, E.H. and Sergy, G.A., 2005. Time Series Observations of Marsh Recovery and Pavement Persistence at Three METULA Spill Sites after 30 1/2 Years. Proceedings 28th Arctic and Marine Oilspill Programme (AMOP) Technical Seminar, Ottawa, ON, 463-472. Owens, Edward H., Mark A. Sienkiewiez, and Gary A. Sergy. "Evaluation of Shoreline Cleaning Versus Natural Recovery: The Metula Spill and Komi Operations." In International Oil Spill Conference, 1999. Page, D. S., P. D. Boehm, J. S. Brown, J. M. Neff, W. A. Burns, and A. E. Bence. "Mussels Document Loss of Bioavailable Polycyclic Aromatic Hydrocarbons and the Return to Baseline Conditions for Oiled Shorelines in Prince William Sound, Alaska." [In eng]. Marine Environmental Research 60, no. 4 (Oct 2005): 422-36. Parker, KR and JA Wiens. 2005. Assessing recovery following environmental : Environmental variation, ecological assumptions, and strategies. Ecological Applications 15:2037-2051. Parker, R. R. Size selective predation among juvenile salmonids fishes in a British Columbia inlet. J. of Fish. Res. Board of Canada, 28:1503–1510 (1971). Parsons, K.C. 1996. Recovering from oil spills: the role of proactive science in mitigating adverse effects. Colonial Waterbirds 19:149-153 Pearson, Leslie. 2006. “M/V Selendang Ayu Response: in Brewer, Reid, Editor. The Selendang Ayu Oil Spill: Lessons Learned. Conference Proceedings, August 16-19, 2005 – Unalaska, Alaska, Available at: www.iccopr.uscg.gov as 2006_Selendang_Ayu_Oil_Spill_Proceeding.pdf Pearson, W.H., D.L. Woodruff, S.L. Kiesser, G.W. Fellingham, and R.A. Elston. 1985. Oil Effects on Spawning Behavior and Reproduction in Pacific Herring (Clupea harengus pallasi). Washington, D.C: American Petroleum Institute. Pearson, W.H., E. Moksness, and J.R. Skalski. 1995. A Field and Laboratory Assessment of Oil Spill Effects on Survival and Reproduction of Pacific Herring Following the Exxon Valdez Spill. In Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters, Edited by P.G. Wells, J.N. Butler and J.S. Hughes, 626-61. Philadelphia: American Society for Testing and Materials. Pearson, W.H., R.A. Elston, R.W. Bienert, A.S. Drum, and L.D. Antrim. 1999. Why Did the Prince William Sound, Alaska, Pacific Herring (Clupea pallasi) Fisheries Collapse in 1993 and 1994?: Review of Hypotheses. Canadian Journal of Fisheries and Aquatic Sciences 56: 711.

July 2012 Page 9-23

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Pearson, W.H., R.B. Deriso, R.A. Elston, S. Hook, K. Parker, and J. Anderson. 2011. Hypotheses Concerning the Decline and Poor Recovery of Pacific Herring in Prince William Sound, Alaska. Reviews in Fish Biology and Fisheries (2011): 1-41. Peterson, C.H. 2001. The Exxon Valdez oil spill in Alaska: acute, indirect and chronic effects on the ecosystem. Advances in Marine Biology. 39: 1-103. Peterson, CH, SD Rice, JW Short, D Esler, JL Bodkin, BE Ballachey, and DB Irons. 2003. Long-term ecosystem response to the Exxon Valdez oil spill. Science 302:2082-2086. Piatt, J.F. and P. Anderson. 1996. Response of Common Murres to the Exxon Valdez oil spill and long- term changes in the Gulf of Alaska marine ecosystem. American Fisheries Society Symposium. 18: 720-737. Piatt, J.F. and R.G. Ford. 1996. How many seabirds were killed by the Exxon Valdez oil spill? American Fisheries Society Symposium.18: 712-719. Piatt, J.F., C.J. Lensink, W. Butler, M. Kendziorek, and D.R. Nysewander. 1990. Immediate impact of the ‘Exxon Valdez’ oil spill on marine birds. The Auk. 107: 387-397. Picou, J. Steven and Cecelia G. Martin. 2006. Long-Term Community Impacts of the Exxon Valdez Oil Spill: Patterns of Social Disruption and Psychological Stress Seventeen Years After the Disaster. Available at: www.arlis.org/docs/vol1/B/243478793.pdf Picou, J. Steven. 2009. “When the Solution Becomes the Problem: The Impacts of Adversarial Litigation on Survivors of the Exxon Valdez Oil Spill” in University of St. Thomas Law Journal. Volume 7, Issue 1. Available at: http://ir.stthomas.edu/ustlj/vol7/iss1/5/ Piper, Ernest and Alaska Department of Environmental Conservation, 1993. The Exxon Valdez Oil Spill, Final Report, State of Alaska Response. Available from NOAA Central Library, Http://www.lib.noaa.gov/ as ExxonValdez_oil_spill_final_report_1993.pdf Plains Midstream Canada. 2012. Clean-up Updates. Available at: http://www.plainsmidstream.com/clean- up Updates Plains Midstream Canada. 2012a. Rainbow pipeline incident – clean-up updates. Available: http://www.plainsmidstream.com/clean-upUpdates. Accessed: 17 June 2012. Plains Midstream Canada. 2012b. Rangeland south pipeline incident: update #9, Sunday, June 17, 2012. Available: http://plainsresponds.files.wordpress.com/2012/06/information-update-_9-june-17- 2012.pdf. Accessed: 17 June 2012. Pontasch, K.W. , and M.A. Brusven. "Diversity and Community Comparison Indices: Assessing Macroinvertebrates Recovery Following a Gasoline Spill." Water Resources 22, no. 5 (1988): 619-26. Prosser, M.V. and Wallace, H.L. (2003) Milford Haven Saltmarsh Survey. A report to the Milford Haven Waterway Environmental Surveillance Group by Ecological Surveys (Bangor). Volume 1 (56 pages + tables) and Volume 2 (Appendices 6 - 8: large format maps).

Page 9-24 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Prosser, M.V. and Wallace, H.L. 2003. Milford Haven Saltmarsh Survey. A report to the Milford Haven Waterway Environmental Surveillance Group by Ecological Surveys (Bangor). Volume 1 (56 pages + tables) and Volume 2 (Appendices 6 - 8: large format maps). Purandare, J.A., T. Huang, M.T. Suidan, and B. Johnston. 1999. Microcosm study of bioremediation of oil-contaminated freshwater wetlands. International Oil Spill Conference. Available: http://www.iosc.org/papers_posters/01251.pdf. Accessed: 17 June 2012. Purcell, J.E., D. Grosse, and J.J. Grover. 1990. Mass Abundance of Abnormal Pacific Herring Larvae at a Spawning Ground in British Columbia. Transactions of the American Fisheries Society 119: 463- 69. Racine, C.H. 1994. Long-term recovery of vegetation on two experimental crude oil spills in interior Alaska black spruce taiga. Canadian Journal of Botany 72(8):1171-1177. Reddy, C.M., Eglinton, T.I., Hounshell, A., White, H.K., Xu, L., Gaines, R.B., and Frysinger, G.S. (2002) The West Falmouth oil spill after thirty years: The persistence of petroleum hydrocarbons in marsh sediments. Env. Sci. and Technol. 36, 4754-4760. Restrepo & Associates. 1982. Ixtoc 1 Oil Spill Economic Impact Study. Prepared for the Bureau of Land Management, U.S. Department of Interior. Available at: http://static.voteig.com.s3.amazonaws.com/files/pdf/lxtoc l 3930.pdf Revesz, Kinga, Tyler B. Coplen, Mary J. Baedecker, Pierre D. Glynn, and Mark Hult. "Methane Production and Consumption Monitored by H and C Isotope Ratios at a Crude Oil Spill Site, Bemindji, Minnesota." Applied Geochemistry 10 (1995): 505-16. Reynolds, J.R. 1979. History and current status of Sitka black-tailed deer in Prince William Sound. Pages 177-183 in Sitka black-tailed deer: Proceedings of a Conference in Juneau Alaska (O.C. Wallmo and J.W. Schoen, eds.). U.S. Department of Agriculture and Forest Service, Juneau, AK. Rice, S. 2008. Significance of Whale Predation on Natural Mortality Rate of Pacific Herring in Prince William Sound. In Exxon Valdez Trustee Council Restoration Project, Progress Report, Project No. 080804, Exxon Valdez Oil Spill Trustee Council. Rice, S. 2009. Significance of Whale Predation on Natural Mortality Rate of Pacific Herring in Prince William Sound. In Exxon Valdez Trustee Council Restoration Project, Progress Report, Project No. PJ090804, Exxon Valdez Oil Spill Trustee Council. Rice, S. D. Effects of oil on fish, pp. 157–182. In: Petroleum effects in the arctic environment (Englehardt, F. R., Ed.) New York: Elsevier Publishers (1985). Rice, S., and M.G. Carls. 2007. Executive Summary. In Prince William Sound Herring: An updated synthesis of population declines and lack of recovery - Final Report, Restoration Project 050794, Edited by S.D. Rice and M.G. Carls. Anchorage: Exxon Valdez Oil Spill Trustee Council. Rice, S.D., R.B. Spies, D.A. Wolfe, and B.A. Wright (eds). 1996. Proceedings of the Exxon Valdez Oil Spill. Symposium. American Fisheries Society Symposium 18, Bethesda, MD.

July 2012 Page 9-25

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Riedman ML, Estes JA. 1990. The sea otter (Enhydra lutris): behavior, ecology, and natural history. US Fish and Wildlife Service Biological Report 90(14). 126 p. Robbins, C.T., J.K. Fortin, K.D. Rode, S.D. Farley, L.A. Shipley, and L.A. Felicetti. 2007. Optimizing protein intake as a foraging strategy to maximize mass gain in an omnivore. Oikos 116: 1675- 1682. Robertson GJ and Goudie RI. 1999. Harlequin Duck (Histrionicus histrionicus), The Birds of North America Online (A. Poole, Ed.). Ithaca: Cornell Lab of Ornithology; Retrieved from the Birds of North America Online: http://bna.birds.cornell.edu/bna/species/466; doi:10.2173/bna.466. Robson, Diana Bizecki, Knight, J. Diane, Farrel, Richard E., Germida, James J. 2004. Natural revegetation of hydrocarbon-contaminated soil in semi-arid grasslands. Canadian Journal of Botany. 82: 22-30. Ross, P.S., and L.S. Birnbaum. 2003. Integrated human and ecological risk assessment: a case study of persistent organic pollutants (POPs) in humans and wildlife. Human Ecol Risk Assess 9:303-24 Rostron, D. (1998) Sea Empress sediment shore impact assessment monitoring: Infauna of heavily oiled shores in Milford Haven and Carmarthen Bay. A report to the Countryside Council for Wales from SubSea Survey, Pembroke. 51pp plus appendices. Rutt G.P., Levell D., Hobbs G., Rostron D.M., Bullimore B., Law R.J. and Robinson A.W. (1998), The effect on the marine benthos. In: Edwards R., Sime H. (Eds.), The Sea Empress Oil Spill: Proceedings of the International Conference held in Cardiff, 11-13 February 1998. Chartered Institute of Water and Environmental Management, London, pp. 189-206. Saba, V. and J.R. Spotila. "Survival and Behavior of Freshwater Turtles after Rehabilitation from an Oil Spill." Environmental Pollution 126, no. 2 (2003): 213-23. Samain J. F., Moal J., Coum A., Le Coz J. R. and Daniel J. Y. (1980) Effects of the "Amoco Cadiz" oil spill on zooplankton. A new possibility of ecophysiological survey. Helgolander Meeresunters. 33, 225-235. Sanchez-Bayo, F., P.J. van den Brink, and R.M. Mann (eds). 2011. Sources and Toxicity of Pollutants. Bentham Books. Sydney, . Sanders SM (2005) Immunotoxicology of Pacific herring: determination of reference ranges and their application to assessing exposure to the water-soluble fraction of crude oil. Dissertation, Simon Fraser University, Burnaby, Bristish Columbia Sanders, H.L.; Grassle, J.F.; Hampson, G.R.; Morse, L.S. and Garner-Price, S. 1981. Long term effects of the barge FLORIDA oil spill, Woods Hole Oceanogr. Inst.,Woods Hole, MA, USA., Report. Schwab, A.P., Banks, K. 1999. Phytoremediation of petroleum-contaminated soils. In: Adriano, D.C., Bollag, J.-M., Frankenburger, W.T.J., Sims. R.C. (Eds.) Bioremediation of Contaminated Soils. American Society of Agronomy, Madison. Seburn, D.C., Kershaw, G.P., Kershaw, L.J. 1996. Vegetation response to a subsurface crude oil spill on a subarctic right-of-way, Tulita (Fort Norman), Northwest Territories, Canada. Arctic 49(4): 321-7.

Page 9-26 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Seeb, J.E., and C. Habicht. 1999. Exxon Valdez Oil Spill Restoration Project Final Report: Laboratory examination of oil-related embryo mortalities that persist in pink salmon populations in Prince William Sound. Restoration Project 97191A-2. Alaska Department of Fish and Game, Anchorage, AK. SEEEC (Sea Empress Environmental Evaluation Committee), (1998). The environmental impact of the Sea Empress oil spill. Final Report of the Sea Empress Environmental Evaluation Committee, 135 pp., London: HMSO. Sell, D., L. Conway, T. Clark, G. B. Picken, J. M. Baker, G. M. Dunnet, A. D. McIntyre, R. B. Clark. (1995) Scientific criteria to optimize oil spill clean-up. Proceedings, 1995 International Oil Spill American Petroleum Institute, Washington, DC. Sellers, R.A., and S.D. Miller. 1999. Population dynamics of brown bears after the Exxon Valdez oil spill. Ursus 11: 73-78. Shackleton, D. 1999. Hoofed mammals of British Columbia. Royal British Columbia Museum Handbook Vol. 3. University of British Columbia Press, Vancouver, BC. Sharp, B.E. 1996. Post-release survival of oil, cleaned seabirds in North America. Ibis 138:222-228. Sharr, S., J. E. Seeb, B. G. Bue, S. D. Moffitt, A. K. Craig, and C. D. Miller. Injury to salmon eggs and pre-emergent fry in Prince William Sound. Restoration study number 60C, final report. 93003 July 1994. Alaska Department of Fish and Game, Anchorage (1994). Short JW, Maselko JM, Lindeberg MR, Harris PM, Rice SD. 2006. Vertical distribution and probability of encountering intertidal Exxon Valdez oil on shorelines of three embayments within Prince William Sound, Alaska. Environmental Science and Technology 40: 3723-3729. Short, J. W., G. V. Irvine, D. H. Mann, J. M. Maselko, J. J. Pella, M. R. Lindeberg, et al. Slightly weathered Exxon Valdez oil persists in Gulf of Alaska beach sediments after 16 years. Envir. Sci. and Tech.41: 1245–1250 (2007). Short, J.W., M.R. Lindeberg, P.M. Harris, J. Maselko, J.J. Pella, and S.D. Rice. 2003. Evaluation of oil remaining in Prince William Sound from the Exxon Valdez oil spill. Final Report, Exxon Valdez Trustee Council Project 02453. EVOSTC, Anchorage, AK. Siciliano, S.D., Germida, J.J. 1998. Mechanisms of phytoremediation: biochemical and ecological interactions between plants and bacteria. Environmental Reviews. 6: 65-79. Simon, T.P. 1999. Assessing the sustainability and biological integrity of water resources using fish communities. CRC Press, Boca Raton, FL. Skalski, J.R., D.A. Coats, and A.K. Fukuyama. 2001. Criteria for oil spill recovery: A case study of the intertidal community of Prince William Sound, Alaska, following the Exxon Valdez oil spill. Environ. Manage. 28:9-18. Sloan, N.A. 1999. Oil impacts on cold-water marine resources : a review relevant to Parks Canada's evolving marine mandate. Parks Canada. National Parks, Occasional paper ; no. 11. 67pp. + iv.

July 2012 Page 9-27

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Smith SDS and Simpson RD (1998) Recovery of benthic communities at Macquarie Island (sub- Antarctic) following a small oil spill. Marine Biology (1998) 131: 567±581. Sparrow, Elena B., Charlotte V. Davenport, and Ronald C. Gordon. "Response of Microrganisms to Hot Crude Oil Spills on a Subarctic Taiga Soil." Arctic 31, no. 3 (September 1978): 324-38. Spies, R.B. 2007. Long-Term Ecological Changes in the Northern Gulf of Alaska. Amsterdam: Elsevier. Spooner, M.F. 1978. Editorial introduction. Amoco Cadiz oil spill. Mar. Pollut. Bull. 9:281-284. Spraker, T.R., L.F. Lowry and K.J. Frost.1994. Gross necropsy and histopathological lesion found in harbor seals. In T.R. Loughlin (ed.). Marine mammals and Exxon Valdez. Academin Press, San Diego, CA. 281-312. St. Aubin, D.J., and J.R. Geraci. 1994. Summary and conclusions. In: Loughlin TR (ed), Marine Mammals and the Exxon Valdez. Academic Press, San Diego, CA, USA, pp 371-376. Stagg R.M., Robinson C., McIntosh A.M., Moffat C.F. and Bruno D.W. (1998). The Effects of the 'Braer' Oil Spill, Shetland Isles, Scotland, on P4501A in Farmed Atlantic Salmon (Salmo salar) and the Common Dab (Limanda limanda). Marine Environmental Research, Volume 46, Number 1, July 1998 , pp. 301-306. Sturdevant, M.V. 1999. Forage Fish Diet Overlap, 1994–1996. Final Report, Restoration Project 97163C. Anchorage: Exxon Valdez Oil Spill Trustee Council. Sublette, K., Jennings, E., Mehta, C., Duncan, K., Brokaw, J., Todd, T., Thoma, G. 2007. Monitoring Soil Ecosystem Recovery Following Bioremediation of a Terrestrial Crude Oil Spill With and Without a Fertilizer Amendment. Soil and Sediment Contamination 16:181-208. Szaro, R.C., N.C. Coon, and W. Stout. 1979. Weathered petroleum: effects on mallard egg hatchability. Journal of Wildlife Management 44:709-713. Tardiff, S.E., and J.A. Stanford. 1998. Grizzly bear digging: effects on subalpine meadow plants in relation to mineral nitrogen availability. Ecology 79: 2219-2228. Teal, J.M., Farrungton, J.W., Burns, K.A., Stegman, J.J., Tripp, B.W., Woodin, B. and Phinney, C. (1992) The West Falmouth oil spill after 20 years: fate of fuel oil compounds and effects on animals. Marine Pollution Bulletin, 24(12), 607-614. Teal, JM and RW Howarth. 1984. Oil Spill Studies: A Review of Ecological Effects. Environmental Management 8:27-44. Templin, W. D., J. S. Collie, and T. J. Quinn, II. Run reconstruction on the wild pink salmon fishery in Prince William Sound,1990–1991. Am. Fish. Soc. Symp. 18: 509–517 (1996). Thorpe, S.A. 1995. Vertical dispersion of oil droplets in strong winds; the Braer oil spill. Marine Pollution Bulletin. 30(11):756-758. Timoney, K.P., and P. Lee. 2009. Does the Alberta tar sands industry pollute? The scientific evidence. Open Conservation Biology Journal 3:65-81.

Page 9-28 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Tourism Economics. 2011, The Impact of the BP Oil Spill on Visitor Spending in Louisiana: Revised estimates based on data through 2010 Q4 http://www.crt.state.la.us/tourism/research/Documents/2011-12/Oil_Spill_Impacts_201106.pdf Trites, A.W. and C.P. Donnelly. 2003. The decline of Steller sea lions Eumetopias jubatus in Alaska: a review of the nutritional stress hypothesis. Mammal Rev. 33(1):3-28. Triton Environmental Consultants Ltd. 2006. Pine River Snorkel Surveys 2005. Prepared for Ministry of Environment, Ft. St. John. Trust, K.A., D. Esler, B.R. Woodin, and J.J. Stegeman. 2000. Cytochrome P450 1A induction in sea ducks inhabiting nearshore areas of Prince William Sound, Alaska. Marine Pollution Bulletin. 40(5): 397-403. U.S. Food and Drug Administration. 2012. Gulf Seafood is Safe to Eat After Oil Spill. Available at: https://blogs.fda.gov/fdavoice/?tag+gulf-seafood U.S. National Pollutions Funds Centre. 2009. Letter to Ms. Margaret Lekanoff, President, Qawalangin Tribe of Unalsaka, RE: Claim Number: J05003-002. Available at: http://www.uscg.mil/npfc/Claims/claims_determinations.asp United States Fish and Wildlife Service (USFWS). 2012. Enbridge oil spill in Michigan’s Kalamazoo River. Available: http://www.fws.gov/midwest/oilspill. Accessed 17 June 2012. US Fish and Wildlife Service (USFWS). 2006. Black-legged Kittiwake: Alaska Seabird Information Series. Available at: http://alaska.fws.gov/mbsp/mbm/seabirds/pdf/blki.pdf. Accessed: February 23, 2012. USA Today. 2011. Tourism returning a year after the Gulf oil spill. Available at: http://www.usatoday.com/news/nation/2011-04-20-tourism-rebounds-gulf-htm USEPA. 2012. Kalamazoo River and Morrow Lake to Open. Available at: http://www.epa.gov/enbridgespill/ Van Hattum, B., C.J.F. Montanes, and C.M.J. Ponds. 1998. Polycyclic aromatic hydrocarbons in freshwater isopods and field-partitioning between abiotic phases. Archives of Environmental Contamination and Toxicology 35:257. Vannote, R.L., G.W. Minshall, K.W. Cummins, J.R. Sedell, and C.E. Cushing. 1980. The River Continuum Concept. Canadian Journal of Fisheries and Aquatic Sciences, 37: 130-137. Wang, H., E. S. Rutherford, H.A. Cook. D.W. Einhouse, R.C. Haas, T.B. Johnson, R. Kenyon, B. Locke, M.W. Turner. 2007. Movement of Walleyes in Lakes Erie and St. Clair Inferred from Tag Return and Fisheries Data. Transactions of the American Fisheries Society. 136: 539 – 551. Wang, Z., Fingas, M. 1997. Developments in the analysis of petroleum hydrocarbon in oils, petroleum products and oil-spill-related environmental samples by gas chromatography. Journal of Chromatography A. 774: 51-78. Wein, R.W., and Bliss, L.C. 1973. Experimental crude oil spills on Arctic plant communities. J. Appl. Ecol. 10: 671 -682.

July 2012 Page 9-29

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Welcomme, R. L., Winemiller, K. O., & Cowx, I. G. (2006). Fish environmental guilds as a tool for assessment of ecological condition of rivers. River Research & Applications, 22(3), 377-396. Wells P.G. (1985) Petroleum hydrocarbons and marine zooplankton. Paper presented to the National Academy of Sciences, Petroleum in the Marine Environment Workshop, November 1981, Clearwater Beach, Florida. Wells, PG, JN Butler, and JS Hughes (eds). 1995. Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters. ASTM STP 1219. ASTM, Philadelphia, PA. Wertheimer AC, Celewycz AG (1996) Abundance and growth of juvenile pink salmon in oiled and non- oiled locations of western Prince William Sound after the Exxon Valdez oil spill. In: Rice SD, Spies RB, Wolfe DA, et al (eds) Proceedings of the Exxon Valdez oil spill symposium, Anchorage, Alaska, February 2-5, 1993, American Fisheries Society Symposium 18, Bethesda, MD, pp 518-532 Whittle, KJ, R Hardy, PR Mackie, and AS McGill. 1982. A quantitative assessment of the sources and fates of petroleum compounds in the marine environment, p 9-34. In: Clark, RB (ed). The Long- Term Effects of Oil Pollution on Marine Populations, Communities, and Ecosystems. The Royal Society. London. Wiens, J.A. 1996. Oil, seabirds and science: The effects of the Exxon Valdez oil spill. BioScience. 46(8): 587-597. Wiens, J.A., R.H. Day, S.M. Murphy and K.R. Parker. 2001. On drawing conclusions nine years after the Exxon Valdez oil spill. The Condor, 103: 886-892. Wiens, J.A., R.H. Day, S.M. Murphy, and K.R. Parker. 2004. Changing habitat and habitat use by birds after the Exxon Valdez oil spill, 1989-2001. Ecological Applications 14: 1806-1825. Wiens, JA, and KR Parker. 1995. Analyzing the effects of accidental environmental impacts: Approaches and assumptions. Ecological Applications 5:1069-1083. Williams, E.H., and T.J. Quinn II. "Pacific Herring, Clupea pallasi, Recruitment in the Bering Sea and North-East Pacific Ocean, I: Relationships among Different Populations." Fisheries Oceanography 9, no. 4 (2000a): 285-99. Williams, E.H., and T.J. Quinn II. "Pacific Herring, Clupea pallasi, Recruitment in the Bering Sea and Wollis, H., and C. Stratmoen. 2010. Population study of western grebes in Alberta 2001-2009: implications for management and status designation. Alberta Species at Risk Report No. 138. Government of Alberta, Sustainable Resource Development. 16 pp. Wooley 2002 In Wooley’s references Woolmer, A.P., Sanderson, W.G., Mackie, A.S.Y., Rees, E.I.S., Hayward, P.J. (in press). The macrobenthos of Carmarthen Bay: an internationally important common scoter (Melanitta nigra) wintering site. Wu, J, and OL Loucks. 1995. From the balance of nature to hierarchical patch dynamics: A paradigm shift in ecology. Quarterly Reviews in Biology 70:439-466.

Page 9-30 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Wyntonyk, D. 2009. Major oil spill at Kinder Morgan B.C. facility. CTV News. Available: http://www.ctvbc.ctv.ca/servlet/an/local/CTVNews/20090507/BC_oil_spill_kinder_morgan_090 507/20090507?hub=BritishColumbiaHome. Accessed 20 June 2012. Ylitalo GM, C.O. Matkin, J. Buzitis, et al. 2001. Influence of life-history parameters on organochlorine concentrations in free-ranging killer whales (Orcinus orca) from Prince William Sound, AK. Sci Total Environ 281:183-203 Zimmerman. S.T., C.S. Gorbics, and L.F. Lowry. 1994. Response activities. In: T.R. Loughlin (ed), Marine Mammals and the Exxon Valdez. Academic Press, San Diego, CA, USA, pp 23-46 Zuberogoitia, I., J.A. Martinez, A. Iraeta, A. Azkona, J. Zabala, B. Jimenez, R. Merino, and G. Gomez. 2006. Short-term effects of the on the peregrine falcon. Marine Pollution Bulletin 52:1176-1181.

Literature Cited for Appendix A and Appendix B Baca, Bart J., Thomas E. Lankford, and Erich R. Gundlach. 1987. Recovery of Brittany Coastal Marshes in the Eight Years Following the Amoco Cadiz Incident. In 1987 International Oil Spill Conference, 6, 1987. Badra, Pete. 2011. Mussel Shell Survey Report: Kalamazoo River Unionid Mussel Shell Survey in the Marshall and Battle Creek Area October 2010. 102. Michigan: Michigan Natural Features Inventory. Bell, B., J. R. Spotila, and J. Congdon. 2006. High Incidence of Deformity in Aquatic Turtles in the John Heinz National Wildlife Refuge. [In eng]. Environmental Pollution 142, no. 3: 457-65. Berkey, Edgar. 1990. Assessment of Environmental Effects from the January 2, 1988 Diesel Oil Spill into the Monogahela River: Final Report on Two Year Study Effort. 288. Pittsburgh, PA: Center for Hazardous Materials Research, University of Pittsburgh, 1990. Boehm, P. D., J. M. Neff, and D. S. Page. 2007. Assessment of Polycyclic Aromatic Hydrocarbon Exposure in the Waters of Prince William Sound after the Exxon Valdez Oil Spill: 1989-2005. [In eng]. Marine Pollution Bulletin 54, no. 3: 339-56. Braddock, Joan F., Jon E. Lindstrom, and Roger C. Prince. 2003. Weathering of a Subarctic Oil Spill over 25 Years: The Caribou-Poker Creeks Research Watershed Experiment. Cold Regions Science and Technology 36, no. 1-3: 11-23. Brewer, Reid. 2005. The Selendang Ayu Oil Spill: Lessons Learned, Conference Proceedings, August 16- 19, 2005, Unalaska, Alaska. Paper presented at the Conference Proceedings August 16-19, 2005- Unalaska, Alaska. Brown, E.D., B.L. Norcross, and J.W. Short. 1996. Introduction to Studies on the Effects of the (Exxon Valdez) Oil Spill on Early Life History Stages of Pacific Herring, (Clupea pallasi), in Prince William Sound, Alaska. Canadian Journal of Fisheries and Aquatic Sciences 53, no. 10: 2337- 42.

July 2012 Page 9-31

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Burk, John C. 1977. A Four Year Analysis of Vegetation Following an Oil Spill in a Freshwater Marsh. Jounal of Applied Ecology 14 (1977): 515-22. Bustard, David, and M. Miles. 2011. Potential Effects of an Oil Pipeline Rupture on Reach 2 of Morice River: A Submission to the Joint Review Panel Enbridge Northern Gateway Project. BC, Canada: Northwest Institute For Bioregional Research Box 2781 Smithers, BC, V0J 2N0 Website: northwestinstitute.ca. Collier, Tracy K., Margaret M. Krahn, Cherl A. Krone, Lyndal L. Johnson, Mark S. Myers, Sin-Lam Chan, and Usha Varanasi. 1993. Oil Exposure and the Effects in Subtidal Fish Following the Exxon Valdez Oil Spill. In 1993 International Oil Spill Conference. Collins, Charles M., Charles H. Racine, and Marianne E. Walsh. 1994. The Physical, Chemical, and Biological Effects of Crude Oil Spills after 15 Years on a Black Spruce Forest, Interior Alaska ARCTIC 47, no. 2: 164-75. Council, The East Walker River Trustee. 2008. Revised Draft Restoration Plan and Environmental Assessment for the Advanced Fuel Filtration Systems East Walker River Oil Spill. Edited by California Department of Fish and Game, Office of Oil Spill Prevention and Response, Nevada Department of Wildlife Fisheries Bureau and Nevada Division of Environmental Protection Bureau of Water Pollution Control. Sacramento, California Reno, Nevada Carson City, Nevada. Cronk, Julie K., William J. Mitsch, and Robert M. Skyes. 1990. Effective Modelling of a Major Inland Oil Spill on the Ohio River. Ecological Modeling 51: 161-92. Cushman, Robert M., and Jonathan C. Goyert. 1984. Effects of a Synthetic Crude Oil on Pond Benthic Insects. Environmental Pollution Series A, no. 33: 163-86. Dauvin, J.C. 1998. The Fine Sand Abra Alba Community of the Bay of Morlaix Twenty Years after the Amoco Cadiz Oil Spill. Marine Pollution Bulletin 36, no. 9: 669-76. de Pennart, H., R. Crowther, T. Taylor, M. Morden, and S. Mattison. 2004. The Use of Ecological Risk Assessment for Regional Management of Aquatic Impacts. Environmental Services Association of Alberta Symposium. Debruyn, Adrian M.H., Barbara G. Wernick, Corey Stefura, Blair G. McDonald, Barri-Lynn Rudolph, Luanne Patterson, and Peter M. Chapman. 2007. In Situ Experimental Assessment of Lake Whitefish Development Following a Freshwater Oil Spill. Environmental Science & Technology 41, no. 20: 6983-89. Driskell, William B., Allan K. Fukuyama, Jonathan P. Houghton, Dennis C. Lees, Gary Shigenaka, and Alan J. Mearns. 1993. Impacts on Intertidal Infauna: Exxon Valdex Oil Spill and Cleanup. In 1993 International Oil Spill Conference. Enbridge Energy. In preparation. 2010/2011 Wildlife Response Report – Fish. Enbridge Line 6b Spill Response. Kalamazoo River, Michigan. Prepared for Enbridge Energy, Limited Partnership. 1601 Pratt Ave., Marshall, Michigan, 49068. Esler, Daniel, and Samuel A. Iverson. 2010. Female Harlequin Duck Winter Survival 11 to 14 Years after the Exxon Valdez Oil Spill. Journal of Wildlife Management 74, no. 3: 471-78.

Page 9-32 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Fukuyama, Allan K., Gary Shigenaka, and Rebecca Z. Hoff. 2000. Effects of Residual Exxon Valdez Oil on Intertidal Protothaca Staminea: Mortality, Growth, and Bioaccumulation of Hydrocarbons in Transplanted Clams. Marine Pollution Bulletin 40, no. 11: 1042-50. Gilfillan, Edward S., Nicole P. Maher, Cecile M. Krejsa, Mary E. Lanphear, Christopher D. Ball, Jeremy B. Meltzer, and David S. Page. 1995. Use of Remote Sensing to Document Changes in Marsh Vegetation Following the Amoco Cadiz Oil Spill (Brittany, France, 1978). Marine Pollution Bulletin 30, no. 12: 780-7. Goldberg, Harry. 2011. Pine River 2011 Fisheries Update: Status of Recovery Post-2000 Pipeline Rupture. Enbridge. Hampton, Steve, Angie Montalvo, Damian Higgins, and Pat Sollberger. 2002. Assessment of Natural Resource Damages as a Result of the East Walker River Oil Spill on December 30 2000. Edited by California Department of Fish and Game - Office of Spill Prevention and Response, U.S. Fish and Wildlife Service and Nevada Division of Wildlife. Hayes, Miles O., and Jacqueline Michel. 1999. Factors Determining the Long Term Persistence of Exxon Valdez Oil in Gravel Beaches. Marine Pollution Bulletin 38, no. 2: 92-101. Hawkins, S.J. and Southward, A.J. 1992. The Torrey canyon oil spill: recovery of rocky shore communities. In: Chapter 13 of Restoring the Nation’s Marine Environment (Thayer, G.W., ed.). pp.583-619. National Oceanic and Atmospheric Administration, Washington D.C. Higgins, Damian K. 2002. Assessment of Damages to Natural Resources in the East Walker River from the Advanced Fuel Filtration Spill: Water, Sediment, and Fish Tissue Analysis. Edited by Interior, 26. Reno, NV: U.S. Fish and Wildlife Services Nevada Fish and Wildlife Office. Hodison, P.V., I. Ibrahim, S. Zambon, A. Ewert, and K. Lee. 2002. Bioavailability of fish to sediment PAH as an indicator of the success of in situ remediation treatments at an experimental oil spill. Bioremediation Journal. 6(3):297 – 313. Hoff, Rebecca Z., Gary Shigenaka, and C. B. Henry, Jr. 1993. Salt Marsh Recovery from a Crude Oil Spill: Vegetation, Oil Weathering, and Response. In 1993 International Oil Spill Conference. Humphrey, B., E. H. Owens, and Gary Sergy. 1991. Long-Term Results from the Bios Shoreline Experiment- Surface Oil Cover. In 1991 International Oil Spill Conference. Irvine, Gail V., Daniel H. Mann, and Jeffrey W. Shorts. 1999. Multi-Year Persistence of Oil Mousse on High Energy Beaches Distant from Exxon Valdez Spill Origin. Marine Pollution Bulletin 38, no. 7: 572-84. Jenkins, T.F., L.A Johnson, Charles M. Collins, and T.T. McFadden. 1978. The Physical, Chemical and Biological Effects of Crude Oil Spills on Black Spruce Forest, Interior Alaska. Arctic 31, no. 3: 305-23. Kingston, Paul F. 2002. Long Term Environmental Impact of Oil Spills. Spill Science & Technology Bulletin 7, no. 1-2: 53-61.

July 2012 Page 9-33

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Kingston, Paul F., I.M.T. Dixon, S. Hamilton, and D.C. Moore. 1995. The Impact of the Braer Oil Spill the Macrobenthic Infauna of the Sediments Off the Shetland Islands. Marine Pollution Bulletin 30, no. 7: 445-59. Kubach, Kevin M., Mark C. Scott, and James S. Bulak. 2011. Recovery of a Temperate Riverine Fish Assemblage from a Major Diesel Oil Spill. Freshwater Biology 56, no. 3: 503-18. Lee, K., R. C. Prince, C. W. Greer, K. G. Doe, J. E. H. Wilson, S. E. Cobanli, G. D. Wohlgeschaffen. 2003. "Composition and Toxicity of Residual Bunker C Fuel Oil in Intertidal Sediments after 30 Years." Spill Science & Technology Bulletin 8, no. 2: 187-99. Lee, Richard F., and David S. Page. 1997. Petroleum Hydrocarbons and Their Effects in Subtidal Regions after Major Oils Spills. Marine Pollution Bulletin 34, no. 11: 928-40. Lee, K. et al. 1999. Natural recovery reduces impact of the 1970 Arrow oil spill. In: Proceedings of the 1999 International Oil Spill Conference. 5pp. Linden, Olof, Ragnar Elmgren, and Paul Boehm. 1979. The Tsesis Oil Spill: Its Impact on the Coastal Ecosystem of the Baltic Sea." Ambio 8, no. 6: 246-53. Lindstrom, Jon E., Ronald P. Barry, and Joan F. Braddock. 1999. Long-Term Ffects on Microbial Communities after a Subarctic Oil Spill. Soil Biology and Biochemistry 31: 1677-89. Lytle, David A., and Barbara L. Peckarsky. 2001. Spatial and Temporal Impacts of a Diesel Fuel Spill on Stream Invertebrates. Freshwater Biology 46: 693-704. MacCarone, Alan D., and John N. Brzorad. 2000. Wading Bird Foraging: Response and Rocovery from an Oil Spill. Waterbirds: The International Journal of Waterbird Biology 23, no. 2: 246-57. MacCarone, Alan D., and John N. Bzrorad. 1998. The Use of Foraging Habitats by Wading Birds Seven Years after the Occurence of Major Oil Spills. Colonial Waterbirds 21, no. 3: 367-74. Michael, A.D., Van Raalte, C.R. and Brown, L.S. 1975. Long-term effects of an oil spill at West Falmouth, Massachusetts. In: 1975 Conference on Prevention and Control of Oil Pollution: Proceedings. Washington, D.C.: American Petroleum Institute. Michel, J., and Myles O. Hayes. 1993. Persistence and Weathering of Exxon Valdez Oil in the Intertidal Zone--3.5 Years Later. In 1993 International Oil Spill Conference, edited by Research Planning Incorporated. Washington, D.C. Michel, J., Z. Nixon, J. Dahlin, D. Betenbaugh, M. White, D. Burton, and S. Turley. 2009. Recovery of Interior Brackish Marshes Seven Years after the Chalk Point Oil Spill. [In eng]. Marine Pollution Bulletin 58, no. 7: 995-1006. Moore, J.J., Taylor, P.M.H. and Hiscock, K. 1995. Rocky shores monitoring programme. In: Monitoring at an Oil Terminal: The Shetland Experience (eds. Dunnet, G.M. and McIntyre, A.D.). Proceedings of the Royal Society of Edinburgh, 103B, 181-200. Moore, J. 2006a. Long term ecological impacts of marine oil spills. In: Proceedings of the Interspill 2006 conference, held at London ExCeL, 21-23 March 2006.

Page 9-34 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Moore, J. 2006b. State of the marine environment in south west Wales 10 years after the Sea Empress oil spill. A report to the Countryside Council for Wales from Coastal Assessment, Liaison & Monitoring, Cosheston, Pembrokeshire. 24pp plus vi. Owens, E. H., B. Humphrey, Dave Hope, Wishart Robson, and John R. Harper. 1987. The Fate of Stranded Oil Four Years after an Experimental Spill on a Sheltered Gravel Beach. In 1987 International Oil Spill Conference. Owens, E. H., E. Taylor, and B. Humphrey. 2008. "The Persistence and Character of Stranded Oil on Coarse-Sediment Beaches." [In eng]. Marine Pollution Bulletin 56, no. 1: 14-26. Owens, Edward H., Mark A. Sienkiewiez, and Gary A. Sergy. 1999. Evaluation of Shoreline Cleaning Versus Natural Recovery: The Metula Spill and Komi Operations. In International Oil Spill Conference. Page, D. S., P. D. Boehm, J. S. Brown, J. M. Neff, W. A. Burns, and A. E. Bence. 2005. Mussels Document Loss of Bioavailable Polycyclic Aromatic Hydrocarbons and the Return to Baseline Conditions for Oiled Shorelines in Prince William Sound, Alaska. [In eng]. Marine Environmental Research 60, no. 4: 422-36. Pontasch, K.W. , and M.A. Brusven. 1988. Diversity and Community Comparison Indices: Assessing Macroinvertebrates Recovery Following a Gasoline Spill. Water Resources 22, no. 5: 619-26. Poulton, Barry C., Edward V. Callahan, Robin D. Hurtubise, and Brad G. Mueller. 1998. Effects of an Oil Spill on Leafpack- Inhabiting Macroinvertebrates in the Chariton River Missouri. Environmental Pollution 99: 115-22. Prince, Roger C., Robert M. Garrett, Richard E. Bare, Matthew J. Grossman, Todd Townsend, Joseph M. Suflita, Kenneth Lee. 2003. "The Roles of Photooxidation and Biodegradation in Long-Term Weathering of Crude and Heavy Fuel Oils." Spill Science & Technology Bulletin 8, no. 2: 145-56. Prince, Roger C., Edward H. Owens, and Gary A. Sergy. 2002. Weathering of an Arctic Oil Spill over 20 Years: The Bios Experiment Revisited. Marine Pollution Bulletin 44, no. 11: 1236. Seymour, Richard J., and Richard A. Geyer. 1992. Fate and Effects of Oil Spills. Annual Reviews of Energy Environment 17: 261-83. Smith SDS & Simpson RD. 1998. Recovery of benthic communities at Macquarie Island (sub-Antarctic) following a small oil spill. Marine Biology 131: 567-581. Sparrow, Elena B., Charlotte V. Davenport, and Ronald C. Gordon. 1978. Response of Microrganisms to Hot Crude Oil Spills on a Subarctic Taiga Soil. Arctic 31, no. 3: 324-38. Spooner, M.F. 1978. Editorial introduction. Amoco Cadiz oil spill. Mar. Pollut. Bull. 9:281-284. Stantec Consulting. In preparation. 2010/2011 Wildlife Response Report – Turtles. Enbridge Line 6b Spill Response. Kalamazoo River, Michigan. Prepared for Enbridge Energy, Limited Partnership. 1601 Pratt Ave., Marshall, Michigan, 49068.

July 2012 Page 9-35

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Section 9: References

Stoker, Sam W. , Jerry Neff, M., Thomas R. Schroeder, and Deborah M. McCormick. 1993. Biological Conditions of Shorelines Following the Exxon Valdex Spill. In 1993 International Oil Spill Conference. Sublette, Kerry L., J. Bryan Tapp, J. Berton Fisher, Eleanor Jennings, Kathleen Duncan, Greg Thoma, Josh Brokaw, and Tim Todd. 2007. Lessons Learned in Remediation and Restoration in the Prairie: A Review. Applied Geochemistry 22, no. 10 : 2225-39. Teal, J.M., Farrungton, J.W., Burns, K.A., Stegman, J.J., Tripp, B.W., Woodin, B. and Phinney, C. 1992. The West Falmouth oil spill after 20 years: fate of fuel oil compounds and effects on animals. Marine Pollution Bulletin, 24(12), 607-614. Van Derveer, William D., Royal J. Nadeau, and Georgia L. Case. 1995. A Screening Level Evaluation of Impacts to a Lotic Macroinvertebrate Community from a Fuel Oil Spill. In Proceedings of 1995 International Oil Spill Conference. Venosa, Albert D., and Xueqing Zhu. 2003. Biodegradation of Crude Oil Contaminating Marine Shorelines and Freshwater Wetlands. Spill Science & Technology Bulletin 8, no. 2: 163-78. Walterhouse. 2012. A Biological Survey of Sites on the Kalamazoo River and Talmadge Creek near the Enbridge Oil Spill in Marshall Calhoun County, Michigan, August 2011. Michigan Department of Environmental Quality Water Resources Division. Wang, Zhendi, Merv Fingas, Sandra Blenkinsopp, Gary Sergy, Michael Landriault, Lise Sigouin, and P. Lambert. 1998. Study of 25-Year-Old Nipisi Oil Spill: Persistence of Oil Residues and Comparisons between Surface and Subsurface Sediments. Edited by ETC Emergencies Science Division, Environment Canada. 3439 River Road, Ottawa, Ontario, Canada, K1A 0H3. Wernick, B. G., A. M. deBruyn, L. Patterson, and P. M. Chapman. 2009. Effects of an Oil Spill on the Regrowth of Emergent Vegetation in a Northern Alberta Lake. [In eng]. Arch Environ Contam Toxicol 57, no. 4: 697-706.

Page 9-36 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

Appendix A Review of Recovery of the Biophysical Environment from Oil Spills

July 2012 Page A-1

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

A.1 Introduction The objective of this report was to review studies that examined ecological recovery from oil spills. Most studies of oil spill recovery focus on one or more Valued Ecosystem Components (VEC) rather than whole ecosystems. The papers that were analyzed were those that documented changes in conditions of a particular VEC with time after an oil spill. Over 114 publications were screened and reviewed for relevance to the Northern Gateway Project. Some of those studies were selected for detailed review and analysis based on similar study area conditions or the other criteria listed below. The results of those focused reviews were presented in the main body of the document. The objective of this Appendix was to present basic descriptive statistics from the broader body of literature compiled and presented in Appendix B. The geographic distribution of these studies is presented in Figure A.1.

July 2012 Page A-3

Attachment 8 to Northern Gateway Reply Evidence

8 40 !( !( !( 36 3 28 12 !( !( !( 11 !( 35 !( 48 !( 39 !(19 !( 43 21 38!( 49 !( !( 32 31!(42 22!(2 !( !( 29 !(!(!( !(!( !( 4 !( 18 23 27 !( 47 1 34 50 14 !( !(!(!( !( 46 !( 6 !(!(!( !( !(!( !( 17 41 20

45 !(!( 45 44

27 15 !( !(24 5 7 25 !( !( !( !( 16 13 !( 30 33 !( 37 !( !( 371313 10 41 16 !(

UID Name UID Name UID Name UID Name UID Name 1 AMOCO CADIZ 12 Caribou-Poker 22 Fidalgo Bay 32 Moose Jaw 42 SEA EMPRESS 2 ARCO ANCHORAGE 13 Chalk Point 23 Flathead Reservation 33 NELLA DAN 43 SELENDANG AYU St. Law rence River Experimental Platform 3 Arctic Island 14 Chariton River 24 Florida 34 Nestucca 44 Wetland Plots !( Barge 4 ARROW 15 Conrail/Cayuga Inlet 25 Heinz Wildlife Refuge 35 Nipisi 45 St. Law rence River Mesocosm !( Experimental !( 5 Arthur Kill 16 Delaw are Mesocosm 27 Kalamazoo River 36 Norman Wells 46 Tallgrass Prairie Preserve Pipeline !( Rail 6 Asher Creek 17 East Walker River 28 Komi Pipeline 37 Patuxent River 47 TORREY CANYON !( Refinery 7 Ashland Spill 18 29 Mackenzie Delta 38 Pine River 48 ERIKA TSESIS !( Storage Tank 8 Baffin Island Experimental 19 ESSO BERNICIA 29 Mackenzie Delta 39 Plains Rainbow 49 Wabamun Lake !( Tanker/Ship Experimental Tennessee !( Truck 10 BAHIA PARAISO 20 Ponds 30 METULA 40 Prudhoe Bay 50 Wolf Lodge Creek 11 BRAER 21 EXXON VALDEZ 31 Milford Haven 41 Reedy River

FigureFigure A.1A.1 OilOil SpillsSpills byby LocationLocation Embridge Northern Gateway Pipelines Project 4370 Dominion Street 5th Floor Burnaby, BC V5G 4L7 Phone 604.436.3014 Fax 604.436.3752 www.stantec.com ´ Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

A.2 Methods Four general criteria were used in selecting oil spill studies for the screening review: 1. Cold Temperature Zone or Subarctic – not tropical or subtropical 2. Recovery information – spill received sufficient study to follow at least some aspect of recovery 3. Similar environment – spill occurred in environment similar to study area 4. Only spill with information Several authors involved in this study possessed extensive personal libraries because of their long history of involvement with oil spills. Literature searches were also conducted using EBSCOhost, ScienceDirect, and Google Scholar. Authors used variations on standard search terms such as “seals”, “oil spills”, and “recovery”. As papers were acquired an analyst reviewed the paper for relevance. If determined suitable to evaluate recovery, basic information was recorded on the nature of the spill. Attributes catalogued included: • The oil spill name • Spill location • Year of spill • Oil type • Volume spilled • Volume recovered in cleanup efforts • Platform of spill (e.g. tanker, pipeline, barge, etc.) • Environment impacted (Marine, Freshwater, Terrestrial) • Valued ecosystem component studied (e.g., fish, aquatic macroinvertebrates, etc.) • Observed mortality • Recovery reported • Processes involved • Measures used to accelerate recovery • Years to recovery • Study duration Reviewers were instructed to accept the conclusions regarding recovery as presented in the paper without assessment of the validity or accuracy of the conclusions. Each paper was evaluated for recovery in the form of quantitative data showing temporal change or simply using general keywords such as “recover (or any part of the word), return to natural equilibrium, return to previous historical conditions, return to pristine conditions, and return to baseline. Reviewers

July 2012 Page A-5

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

evaluated the study under consideration to ensure that adequate information existed to draw conclusions about recovery status (e.g., the use of baselines or of undisturbed reference versus control areas). The status of recovery was grouped into five main categories, based on the study author’s presentation of recovery in each paper. • None or none apparent = not showing recovery, long term effects have set in and conditions are not improving or improve very modestly; • Recovered = full recovery or a return to pre-spill conditions that was equal to non-oil spill base data or baseline conditions; • Recovering = partial recovery, moving closer to conditions before oil spill, but not yet exhibiting full recovery; • Uncertain = this category refers to circumstances where some metrics, sample locations, or time periods suggest that some recovery has occurred, while other indicators suggest that recovery has not occurred; • Insufficient Data = the author mentions recovery but did not present sufficient information or data for the reader to decipher recovery Once the data base had been assembled it was possible to generate simple descriptive statistics characterizing the type and number of spills encountered in the literature, the duration of study by ecosystem, the number and volume of spills by ecosystem and other appropriate indicators. Because of the breadth and number of studies reviewed, the data summaries provide a useful characterization of oil spills and recovery in temperate ecosystems. For detailed review, a total of 50 oil spills were selected from the original 114 studies. (Appendix B). The oil spills were segregated into four environment categories: marine, freshwater, terrestrial, and terrestrial/freshwater. The platform was defined as that the vessel, storage unit, transmission line, or other device that contained the oil prior to the spill. The volume of spilled oil was converted from the reported units (e.g., liters, gallons, etc.) to tonnes (metric tons) for all spills. Most of the studies focused on the response of individual organisms or groups of organisms when evaluating recovery. For this analysis it was necessary to consolidate species into generalized groups. The VECs were classified into 12 general categories as presented below: • Algae • Birds • Fish • Macroinvertebrates • Mammals • Microbial Community

Page A-6 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

• Reptiles • Vegetation • Sediment • Shoreline • Soil • Water Quality

A.3 Results Pipelines (N= 15) and Tankers/Ships (N= 14) jointly accounted for 60% of the total spills while 22% were experimental spills (N=11) (Table A.1). The experimental spills refer to oil that was intentionally released into the environment for the purpose of studying its behavior. Barges accounted for five additional spills, and spills associated with railways totaled two. Mean spill volume was greatest for tankers/ships (47,415 tons) and pipelines (11,765). The standard deviations for these categories are large indicating considerable variability and large outliers represented in the data by a single spill (e.g., Amoco Cadiz –tanker—240,000 metric tons). Mean study duration ranged between 2.5 and 15.2 years. Some of the experimental studies have lasted 20-25 years (e.g. Caribou-Poker). In general it seems that small spill volumes are studied for shorter duration than larger spills.

Table A.1 Oil spill frequencies by platform, volume spilled and study duration. Standard Mean Oil Deviation Study Standard Deviation Platform Spills Mean Volume Volume Duration Study Duration (metric tons) (metric tons) (years) (years) Barge 2 646 89 15.2 14.8 Experimental 11 10,607 27948 9.6 9.2 Pipeline 15 11,765 34626 5.6 7.1 Rail 2 76 53 2.5 3.4 Refinery 2 670 No data 5.9 7.1 Storage Tank 1 2,234 No data 1.8 0.6 Tanker/Ship 14 47,415 68,616 9.0 8.0 Truck 1 11 No data 2.9 2.7 Total 48 21,304 47,695 7.1 8.9 Marine (N= 24) and freshwater spills (N= 15) account for 81 percent of total number of spills (Table A.2). The mean volume for marine spills was highest at approximately 28,204 metric tons. The other classes were far lower with means ranging between 704 to 13,377 metric tons. The total mean volume in the freshwater environment is relatively small, despite having 15 spills under investigation. Marine case studies in the 10,001 – 100,000 metric tons range are inversely

July 2012 Page A-7

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

proportional to the VEC count for this category (Figure A.2 & A.3). This is due to the high number of VEC’s studied for Exxon Valdez which was also a high volume spill.

Table A.2 Oil spill frequencies by environment, volume spilled and study duration. Standard Standard Mean Deviation Oil Deviation Study Study Environment Spills Mean Volume Volume Duration Duration (metric tons) (metric tons) (years) (years) Freshwater 15 13,376.9 38911.2 1.6 1.9 Marine 24 28,203.5 55,345.9 8.5 8.3 Terrestrial 8 12,371.5 27409.6 15.9 9.3 Terrestrial, Freshwater 1 704.0 No Data 6.5 4.0 Total 48

Figure A.2 Comparison of the volume (mean metric tons) and frequency of oil spills in various environments. Spill volume not available (No Data; N= 5)

Page A-8 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

Figure A.3 Comparison of oil spill volume (metric tons) and VEC count by environment. Spill volume not available (No Data; N = 6) In general, the majority of freshwater spill VEC’s were studied less than 2 years and the marine spill VEC’s were studied for more lengthy periods (Figure A.4). There were 5 terrestrial VEC’s studies that were studied for more than 20 years. These were experimental spill studied VEC’s for Caribou-Poker and Mackenzie Delta.

Figure A.4 Comparison of study duration (years) for four environment types based on the VEC count. (Unknown Study Duration; N=33)

July 2012 Page A-9

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

A.3.1 Valued Ecological Component (VEC): Based on the 48 oil spills and 98 publications, a total of 174 different VECs were analyzed for recovery status from the reviewed literature (Table A.3). Macroinvertebrates were the most frequently studied group which included a wide range of fauna including benthic epifauna, benthic infauna, amphipods, insects, shellfish and coral. The abiotic categories referred to the oil located in sediment or part of the soil profile. Shoreline was the most general term used which included oiled beaches, and oil located at the intertidal or subtidal zones.

Table A.3 Valued Ecological Components (VEC) associated with 48 oil spills. VEC Total Count Marine Freshwater Terrestrial F & T Biotic Algae 5 5 0 0 0 Birds 11 7 4 0 0 Fish 21 9 10 0 2 Macroinvertebrates 46 33 12 0 1 Mammals 3 0 3 0 0 Microbial Community 4 1 1 2 0 Reptiles 2 0 0 2 0 Vegetation 17 5 4 8 0 Abiotic Sediment 12 7 4 0 1 Shoreline 36 35 1 0 0 Soil 13 5 0 8 0 Water Quality 4 2 2 0 0 Total 174 109 43 18 4

A.3.2 Recovery Status: There were a total 140 VECs (not including the Insufficient Data or Uncertain) for which the author(s) clearly addressed recovery over time (Table A.4). These VECs will be the focus of the remainder of discussion. In most cases, the studies have a clearly defined study duration or time after spill. For the VECs studied, the number of those that show no sign of recovery (None or none apparent) or partial recovery (Recovering) gradually decreases with increasing study duration (Figure A.5). The VEC’s that experienced “Recovery” happened most frequently in the 5-10 year range (Figure A.5).

Page A-10 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

Table A.4 Recovery status for Valued Ecological Components (VECs) based on mean and standard deviation of study durations. Standard Deviation Recovery Status VEC Count Mean Study Duration Study Duration (years) (years) None or none apparent 28 7.82 8.68 Recovered 61 8.24 7.03 Recovering 51 7.37 8.61 Uncertain 12 7.26 11.36 Insufficient Data 22 3.63 6.38 Total 174 7.12 8.22

Figure A.5 Frequency of Valued Ecological Components (VECs) and recovery status based on study duration. There were 32 VECs with an unknown study duration.

July 2012 Page A-11

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

One hundred and twelve VECs of the 174 examined were classified as recovered or recovering and 28 VECs showed no sign of recovery. In 22 cases, reviewers felt that data were insufficient to adequately judge recovery status, and in 12 VECs the study authors were uncertain regarding the recovery status. Based on the average years to recover we evaluated the mean time after the spill and the mean time to recover for applicable VECs (Table A.5). We included those studies that indicated recovery so that the mean time to recover for four environment types could be standardized for comparison. Over half of the marine studies showed recovery in 5.6 years, while 50% of the studies (n=16) in terrestrial systems showed no sign of recovery in 16 years (Table A.5). Freshwater system studies showed over 25% full recovery in 2.3 years after spills, while the limited information for terrestrial and freshwater environments showed that between 4.5 to 8.5 years after the spill that about a half attained pre-spill recovery or were still recovering, respectively (Table A.5).

Table A.5 Recovery status by environment showing the mean time to recover post spill and the mean time to recover in years. Mean Study Mean Time to Environment Recovery Status VEC Count Duration Recovered (years) (years) Freshwater None or none apparent 7 2.13 N/A Recovered 6 3.46 2.28 Recovering 10 0.99 N/A Total 23 Marine None or none apparent 13 5.50 N/A Recovered 51 9.29 5.57 Recovering 33 8.01 N/A Total 97 Terrestrial None or none apparent 8 16.00 N/A Recovered 2 2.30 2.00 Recovering 6 25.00 N/A Total 16 Terrestrial/Fresh- None or none apparent 0 N/A N/A water Recovered 2 4.50 3.00 Recovering 2 8.50 N/A Total 4

Page A-12 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

There were 12 different categories for types of oil spilled (Figure A.6). Half of the spill categories involved two or more types of oil. Crude oil and light crude oil were the most frequently reported and studied spills in the data base. It is not possible to identify any clear patterns regarding recovery due to the paucity of data for certain oil types.

Figure A.6 Recovery status by type of oil spilled.

A.4 Conclusion This review examined 48 individual oil spills and focused on 174 Valued Ecosystem Components. These studies examined marine, freshwater, and terrestrial oil spills originating from a variety of sources. The bullets below provide the major findings and conclusions from this review. • Tankers, pipelines, and experimental spills accounted for 83% of the spills reviewed. • Mean spill volume was greatest for tankers/ships (47,415 metric tons). • Mean study duration for spills was 7.1 years. • Marine spills were larger on average (28,204 metric tons), than freshwater (13,377 metric tons), or terrestrial (12,372 metric tons) spills. • Spills between 101 and 1,000 metric tons were most frequently studied and reported. • VEC counts were highest for large marine spills • Relatively few freshwater VECs were studied for longer than two years while marine VECs were frequently studied for five to ten years.

July 2012 Page A-13

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix A: Review of Recovery of the Biophysical Environment from Oil Spills

• Aquatic macroinvertebrates (freshwater and marine) were the most frequently studied biotic group and shorelines were the most frequently studied abiotic group • Recovery was observed on average after 5.1 years. • Freshwater spills were most frequently characterized as recovering whereas marine spills were most frequently reported as recovered. • Crude and light crude oil were the most frequently reported spills although spills were often reported for mixtures of two or more kinds of product.

A.5 Summary These data provide useful perspective on oil spills and ecological recovery. Marine spills appear to be relatively well studied and much is known about the rate of recovery (mean = 5.8 years). Freshwater and terrestrial spills have been less intensively studied. In freshwater systems recovery happened quickly when it was observed (mean = 2.3 years).

Page A-14 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix B: List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills

Appendix B List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills

July 2012 Page B-1

Attachment 8 to Northern Gateway Reply Evidence

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix B: List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills

Table B.1 List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills Volume Oil Spill (metric Environmental Years to Study UID Name Location Year Oil Type tons) Platform Environments Component Studied Recovery Status recover Duration References 1 AMOCO Brittany, 1978 Light crude 240,000 Tanker/Ship Marine Macroinvertebrates Recovered 6 18 Kingston 2002 CADIZ France Macroinvertebrates Recovered 10 18 Kingston 2002 Macroinvertebrates Recovered 2 18 Kingston 2002 Shoreline Recovered 8 18 Kingston 2002 Vegetation Recovered 4.5 5 Baca et al. 1987 Macroinvertebrates Recovered 15 20 Dauvin 1998 Macroinvertebrates Uncertain N/A 20 Dauvin 2000 Vegetation Recovered 4.5 13 Gilfillan et al. 1995 Macroinvertebrates Recovered 2 6 Seymour and Geyer, 1992 Macroinvertebrates Recovered 10 Unknown Seymour and Geyer, 1992 Macroinvertebrates Recovered 1 3 Seymour and Geyer, 1992 Macroinvertebrates Recovered 2.5 3 Seymour and Geyer, 1992 Fish Recovered 0.5 3 Seymour and Geyer, 1992 Macroinvertebrates Recovered 10 Unknown Lee and Page 1997 Shoreline Recovering N/A 8 Baca et al 1987 Algae Recovered 0.25 Unknown Spooner 1978 2 ARCO Port Angeles, 1985 Crude 776 Tanker/Ship Marine Shoreline Recovered 1 Unknown Pearson et al. 1986 ANCHORAGE WA 3 Arctic Island Baffin Island 1981 Crude 13 Experimental Marine Shoreline Recovering N/A 3.28 Lee and Page 1997 4 Arrow Chedabucto 1970 No. 6 Fuel Oil 7,980 Tanker/Ship Marine Shoreline Recovered 30 Unknown Owens 1978 Bay, Nova Macroinvertebrates Recovered 1.5 Unknown Conover 1971 Scotia, CA Shoreline None or none apparent N/A 20 Kingston 2002 Macroinvertebrates Recovered 6.5 13 Gilfillian et al. 1977 Soil Insufficient Data N/A 4 Prince et al. 2003 Shoreline Insufficient Data N/A 4 Prince et al. 2003 Macroinvertebrates Recovering N/A 7 Lee et al. 2003 Fish Recovering N/A 7 Lee et al. 2003 Microbial Community Recovering N/A 7 Lee et al. 2003 Soil Insufficient Data N/A 30 Owens et al. 2008 Shoreline Recovering N/A 30 Owens et al. 2008 Shoreline Recovered 7 Unknown Thomas 1978 Shoreline Uncertain N/A 27 Lee et al 1999

July 2012 Page B-3

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix B: List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills

Table B.1 List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills (cont’d) Volume Oil Spill (metric Environmental Years to Study UID Name Location Year Oil Type tons) Platform Environments Component Studied Recovery Status recover Duration References 5 Arthur Kill NY,NJ, USA 1990 No. 6 Fuel Oil & 4,216 Pipeline Marine Birds Recovering N/A 1 Maccarone and Brzorad No. 2 Fuel Oil 1998 Birds Recovering N/A 1 Maccarone and Brzorad 2000 6 Asher Creek Missouri, US 1979 Crude 1,000 Pipeline Freshwater Macroinvertebrates Recovered 1.5 1.5 Seymour and Geyer 1992 7 Ashland Spill Monongahela 1988 No. 2 Fuel Oil 2,234 Tanker/Ship Freshwater Macroinvertebrates Insufficient Data N/A 2 Berkey et al. 1990 River, PA, US Vegetation Insufficient Data N/A 2 Berkey et al. 1990 Fish Insufficient Data N/A 2 Berkey et al. 1990 Birds Insufficient Data N/A 2 Berkey et al. 1990 Mammals Insufficient Data N/A 2 Berkey et al. 1990 Microbial Community Insufficient Data N/A 2 Berkey et al. 1990 Sediment Uncertain N/A 2 Berkey et al. 1990 Sediment Recovering N/A 0.083 Cronk et al. 1990 8 Baffin Island Cape Hatt, 1981 Crude 15 Experimental Marine Soil Recovering N/A 20 Prince et al. 2002, 2003 Experimental Baffin Island, Shoreline Recovering N/A 20 Prince et al. 2002, 2003 Spill Nunavut, CA Shoreline Recovering N/A 8 Owens et al. 2008 Shoreline Recovering N/A 4 Owens et al. 1987 Shoreline Recovering N/A 9 Humphrey et al. 1991 10 Bahia Paraiso 1989 No. 2 Fuel Oil 500 Tanker/Ship Marine Macroinvertebrates Recovering N/A 2 Seymour and Geyer, 1992 11 BRAER , UK 1993 Crude & No. 6 Fuel 84,000/1,475 Tanker/Ship Marine Shoreline Recovered Unknown Unknown Kingston 2002 Oil Fish Recovered 0.17 2 Kingston. 1999 Shoreline Recovered 1 2 Kingston. 1999 Macroinvertebrates Recovered 1 2 Kingston. 1999 Macroinvertebrates Recovering N/A 2 Kingston. 1999 Macroinvertebrates None or none apparent N/A 0.41 Kingston et al. 1995 12 Caribou-Poker Fairbanks, 1976 Crude 7 Experimental Terrestrial Soil None or none apparent N/A 25 Braddock et al. 2003 Alaska Soil None or none apparent N/A 15 Collins et al. 1994 Vegetation None or none apparent N/A 15 Collins et al. 1994 Soil None or none apparent N/A 25 Prince et al. 2003 Vegetation None or none apparent N/A 25 Prince et al. 2003 Soil Insufficient Data N/A 10 Sparrow and Sparrow. 1987 Microbial Community None or none apparent N/A 2 Sparrow et al. 1978 Vegetation Insufficient Data N/A Unknown Racine. 1994 Microbial Community None or none apparent N/A 19 Lindstrom et al. 1999 Vegetation None or none apparent N/A 2 Jenkins et al. 1978

July 2012 Page B-5

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix B: List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills

Table B.1 List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills (cont’d) Volume Oil Spill (metric Environmental Years to Study UID Name Location Year Oil Type tons) Platform Environments Component Studied Recovery Status recover Duration References 13 Chalk Point , US 2000 No. 6 Fuel Oil & 447 Pipeline Marine Soil None or none apparent N/A 7 Michel et al. 2009 No. 2 Fuel Oil Vegetation Recovering N/A 7 Michel et al. 2009 Macroinvertebrates None or none apparent N/A 7 Michel et al. 2009 14 Chariton River Chariton River, 1990 Light crude 356 Pipeline Freshwater Macroinvertebrates Recovering N/A 0.16 Poulton et al. 1998 Ethel, Missouri, US 15 Conrail/Cayug West Danby, 1997 No. 2 Fuel Oil 23 Rail Freshwater Macroinvertebrates Recovering N/A 1.25 Lytle and Peckarsky 2001 a Inlet New York, US 16 Delaware Delaware Bay, 1994 Crude 0.1 Experimental Marine Shoreline None or none apparent N/A 1.23 Venosa and Zhu 2003 Mesocosm Dover, Delaware 17 East Walker California/ 2000 No. 6 Fuel Oil 11 Truck Freshwater Fish Uncertain N/A 0.21 Higgins et al. 2002 River Nevada Sediment Uncertain N/A 0.41 Higgins et al. 2002 Water Quality Uncertain N/A 0.41 Higgins et al. 2002 Fish Uncertain N/A 0.21 Higgins et al. 2002 Fish Insufficient Data N/A 1 East Walker River Trustee Council 2008 Birds Insufficient Data N/A 1 East Walker River Trustee Council 2008 Mammals Insufficient Data N/A 1 East Walker River Trustee Council 2008 Fish Insufficient Data N/A 1 Hampton et al, 2002 Birds Insufficient Data N/A 1 Hampton et al, 2002 Mammals Insufficient Data N/A 1 Hampton et al, 2002 Macroinvertebrates Recovering N/A 1 Hampton et al, 2002 Macroinvertebrates Insufficient Data N/A 1 Hampton et al, 2002 18 Erika Brittany, 1999 No. 6 Fuel Oil 28,000 Tanker/Ship Marine Shoreline Recovering N/A 0.3 Prince et al. 2003 France Soil Recovering N/A 0.3 Prince et al. 2003 19 Esso Bernicia Sullom Voe, 1978 No. 6 Fuel Oil 1,174 Tanker/Ship Marine Shoreline Recovered 15 34 Moore et al. 1995 Shetland, UK 20 Experimental Tennessee, US 1981 Crude No Data Experimental Freshwater Macroinvertebrates Recovering N/A 0.91 Cushman and Goyart, Tennessee 1984 Ponds

July 2012 Page B-7

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix B: List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills

Table B.1 List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills (cont’d) Volume Oil Spill (metric Environmental Years to Study UID Name Location Year Oil Type tons) Platform Environments Component Studied Recovery Status recover Duration References 21 EXXON Prince William 1989 Crude 33,000 Tanker/Ship Marine Shoreline Recovering N/A 6 Owens et al. 2008 VALDEZ Sound, Alaska Birds Recovered Unknown Unknown Kingston 2002 Birds None or none apparent N/A 9 Kingston 2002 Algae Recovered 3 8 Kingston 2002 Shoreline Recovered 4 8 Kingston 2002 Birds Recovered 10 13 Esler and Iverson 2008 Birds Insufficient Data N/A 4 Esler et al. 2010 Shoreline Recovering N/A 3 Michel and Hayes 1993 Shoreline Recovering N/A 3 Stoker et al. 1993 Fish Insufficient Data N/A 3 Collier et al. 1993 Macroinvertebrates Recovering N/A 2 Driskell et al. 1993 Water Quality Recovering N/A 6 Boehm et al. 2007. Macroinvertebrates Recovering N/A 13 Boehm et al. 2007. Macroinvertebrates Uncertain N/A 3 Fukuyama et al. 2000 Sediment Recovered 3 8 Hayes and Michel. 1999 Sediment Recovered 2 8 Hayes and Michel. 1999 Sediment None or none apparent N/A 8 Hayes and Michel. 1999 Macroinvertebrates Recovering N/A 8 Hoff and Shigenaka 1999 Macroinvertebrates Recovered 2 8 Hoff and Shigenaka 1999 Macroinvertebrates Recovered 4 8 Hoff and Shigenaka 1999 Water Quality Recovered 4 8 Hoff and Shigenaka 1999 Algae Recovered 4 8 Hoff and Shigenaka 1999 Sediment Recovered 0.25 8 Hoff and Shigenaka 1999 Sediment Recovered 6 8 Hoff and Shigenaka 1999 Shoreline None or none apparent N/A 4.9 Irvine et al. 1999 Sediment Recovered 2 Unknown Lee and Page 1997 Macroinvertebrates Uncertain N/A 3 Lee and Page 1997 Fish Recovered 2 Unknown Lee and Page 1997 Fish Recovered 1 2 Pearson et al. 1995 Fish Recovered 2 2 Brown et al. 1996 Macroinvertebrates None or none apparent N/A 1 Lee and Page 1997 Shoreline Recovering N/A 3.28 Lee and Page 1997 Macroinvertebrates Recovered 11 12 Page et al. 2005 Shoreline Recovered 0.6 Unknown Lee and Page 1997

July 2012 Page B-9

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix B: List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills

Table B.1 List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills (cont’d) Volume Oil Spill (metric Environmental Years to Study UID Name Location Year Oil Type tons) Platform Environments Component Studied Recovery Status recover Duration References 22 Fidalgo Bay Puget Sound, 1991 Crude 670 Refinery Marine Vegetation Recovering N/A 1.25 Hoff et al. 1993 WA Sediment Uncertain N/A 0.5 Hoff et al. 1993 23 Flathead Montana 1993 Multiple No Data Pipeline Freshwater Macroinvertebrates Recovering N/A 1 Van derVeer et al. 1994 Reservation 24 FLORIDA Falmouth, MA, 1969 No. 2 Fuel Oil 557 Barge Marine Shoreline Recovered 12 Unknown Sanders & Teal USA Macroinvertebrates Recovered 20 Unknown Sanders & Teal Shoreline Uncertain N/A 30 Teal et al 1992 Macroinvertebrates Recovering 5 Unknown Michael et al 1975 25 Heinz Wildlife Darby Creek, 2000 Crude 650 Pipeline Freshwater Reptiles None or none apparent N/A 3 Bell et al. 2006 Refuge Pennsylvania 27 Kalamazoo Michigan, US 2010 Heavy crude 3,247 Pipeline Freshwater Macroinvertebrates None or none apparent N/A 0.25 Badra 2011 River Reptiles Recovered Unknown Unknown Enbridge, Stantec In Prep Fish Recovering N/A Unknown Enbridge In Prep Birds Recovered Unknown Unknown Enbridge, Stantec In Prep Macroinvertebrates None or none apparent N/A 0.083 Badra 2010 Macroinvertebrates Recovering N/A 2 Walterhouse 2012 28 Komi pipeline 1995 Crude 136,370 Pipeline Freshwater Shoreline Insufficient Data N/A 2 Owens et al. 1999 29 Mackenzie Mackenzie 1972 Crude 266 Experimental Terrestrial Soil Recovering N/A Unknown Ross et al., 1973 Delta Delta, Canada Vegetation Recovering N/A Unknown Ross et al., 1973 30 METULA Puerto Espora 1974 Light crude 52,000 Tanker/Ship Marine Shoreline Recovering N/A 36 Owens et al. 2008 Shoreline None or none apparent N/A Unknown Kingston 2002 Shoreline Recovering N/A 10 Owens et al. 1999 31 Milford Haven Milford Haven, 1969 No. 6 Fuel Oil No Data Refinery Marine Shoreline Recovering N/A 16 Baker et al 1993 UK 32 Moose Jaw Moose Jaw, 1974 Crude 2,143 Pipeline Terrestrial Vegetation Recovering N/A Unknown De Jong, 1980 Saskatchewan 33 Nella Dan Macquarie 1987 No. 2 Fuel Oil & 120/5 Tanker/Ship Marine Macroinvertebrates Recovering N/A Unknown Smith & Simpson 1998 Island, Lubricating Oil Antarctic 34 NESTUCCA OR, WA, BC 1988 No. 6 Fuel Oil 734 Barge Marine Fish Recovered 1 1 DFO 2012 Fish Uncertain N/A 0.4 HayandMcCarter, 1995 35 Nipisi Alberta, 1970 - 1972 Crude 8,182 Pipeline Terrestrial Soil Recovering N/A 25 Wang et al. 1998 Canada 36 Norman Wells Norman Wells, 1972 Crude 7 Experimental Terrestrial Vegetation Recovering N/A Unknown Hutchinson and N.W.T. Freeedman, 1978

July 2012 Page B-11

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix B: List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills

Table B.1 List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills (cont’d) Volume Oil Spill (metric Environmental Years to Study UID Name Location Year Oil Type tons) Platform Environments Component Studied Recovery Status recover Duration References 37 Patuxent River Chesapeake 2000 Crude 447 Pipeline Marine Macroinvertebrates None or none apparent N/A 0.5 Versar and ENTRIX 2001 Bay, Maryland, USA 38 Pine River British 2000 Light crude 704 Pipeline Terrestrial, Fish Recovered 5 8 Bustard and Miles. 2011 Columbia, Freshwater Sediment Recovering N/A 5 Bustard and Miles. 2011 Canada Macroinvertebrates Recovered 1 1 De Pennart 2011 Fish Recovering N/A 12 Goldberg. 2011. 39 Plains Alberta, 2011 Crude 3,818 Pipeline Terrestrial Soil Recovering N/A 25 Wang et al. 1998 Rainbow Canada 40 Prudhoe Bay Pudhoe Bay, 1976 Crude & No. 2 Fuel 0.01 Experimental Terrestrial Vegetation Recovered Unknown Unknown Walker et al., 1978 and Alaska Oil McKendrick and Mitchel, 1978 41 Reedy River Reedy River, 1996 No. 2 Fuel Oil 3,057 Pipeline Freshwater Fish Recovered 4 9 Kubach et al. 2011 South Carolina, US 42 Sea Empress Milford Haven, 1996 Crude & No. 6 Fuel 72,000/480 Tanker/Ship Marine Shoreline Recovered 5 Unknown Moore 2006 UK Oil Macroinvertebrates Recovered 5 Unknown Moore 2006 Algae Recovered 1 Unknown Moore 2006 43 Selendang Dutch Harbor, 2004 No. 6 Fuel Oil 1,072 Tanker/Ship Marine Birds None or none apparent N/A 1.5 Brewer et al. 2005 Ayu Alaska 44 St. Lawrence Ste Croix de 1999 Crude No Data Experimental Freshwater Vegetation Recovering N/A 1.25 Hodson, et al. 2002 River Lotbiniere, Fish Recovering N/A 1.25 Hodson, et al. 2002 Experimental Quebec Wetland Plots 45 St. Lawrence St. Lawrence 1999 Light crude No Data Experimental Freshwater Vegetation None or none apparent N/A 0.4 Venosa and Zhu 2003 River River, Ste. Mesocosm Croix, Quebec, Canada 46 Tallgrass Oklahoma, US 1993 - 2002 Crude 84,549 Experimental Terrestrial Soil Recovered 2 2.3 Sublette et al. 2007 Prairie Preserce 47 TORREY Cornwall, UK 1967 Light crude 118,000 Tanker/Ship Marine Vegetation Recovering N/A 4 Burk 1977 CANYON Shoreline Recovered 15 Unknown Hawkins and Southward 1992 48 Tsesis Baltic 1977 No. 5 Fuel Oil & 1,000 Tanker/Ship Marine Macroinvertebrates None or none apparent N/A Unknown Kingston 2002 No. 6 Fuel Oil Shoreline Recovered 1 Unknown Linden et al 1979 Algae Recovered 0.1 Unknown Johansson et al 1980

July 2012 Page B-13

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix B: List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills

Table B.1 List of Oil Spills in Review of Recovery of the Biophysical Environment from Oil Spills (cont’d) Volume Oil Spill (metric Environmental Years to Study UID Name Location Year Oil Type tons) Platform Environments Component Studied Recovery Status recover Duration References 49 Wabamun Alberta, 2005 No. 6 Fuel Oil 126 Rail Freshwater Vegetation Recovered Unknown 2 Wernick et. al 2009 Lake Canada Water Quality None or none apparent N/A 0.16 Anderson, 2006 Fish None or none apparent N/A 1 Transport Safety Board of Canada, 2005 Sediment None or none apparent N/A 10 Zrum and Sergy, 2006 Fish Insufficient Data N/A 0.416 Debruyn et al. 2007 50 Wolf Lodge Idaho, US 1983 Gasoline 72 Pipeline Freshwater Macroinvertebrates Recovered 1.33 1.33 Pontasch and Brusven, Creek 1988

July 2012 Page B-15

Attachment 8 to Northern Gateway Reply Evidence

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

Appendix C Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

July 2012 Page C-1

Attachment 8 to Northern Gateway Reply Evidence Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

C.1 Introduction In written and oral evidence before the Joint Review Panel (JRP) for the Enbridge Northern Gateway Pipelines Project (NGP), intervenors have focused on the Exxon Valdez oil spill (EVOS) as their primary “model of choice” for evaluating potential effects on the human environment, including effects on Aboriginal people and coastal communities (including non- Aboriginal people); specifically: • Heiltsuk Tribal Council Submission of the Potential Effects of an Oil Spill on the Heiltsuk (A38059) • Gitga’at First Nations Submission A Social Impact Assessment of the Enbridge Northern Gateway Pipeline Project in Regard to the Gitga’at First Nation (A2K4W8) • Being Gitga’at: A Baseline Report (A2K4X3) • United Fishermen and Allied Workers Union Potential Impacts of the Enbridge Northern Gateway Pipeline Project on Members of the United Fishermen and Allied Workers’ Union (Liesel Ashley Ritchie and Duane A Gill) The above noted intervenors claim that there will inevitably be a catastrophic marine oil spill, and that this will result in permanent social, cultural and ecological damage to coastal First Nations and coastal communities and their way of life. In reply, this review examines the types of effects and major lessons learned from two Alaskan experiences: the Exxon Valdez oil spill (EVOS) in March 1989 and the Selendang Ayu oil spill (SAOS) in December 2004, more than seventeen years later.

C.2 Background This review compares and contrasts the way the EVOS and SAOS affected the human environment. Fundamental changes in the way that oil spill emergency response in Alaska and elsewhere are now conducted help to mitigate effects on the human environment. The more recent Selendang Ayu spill is an example of how industry, government and communities have changed the paradigm for oil spill mitigation, response and recovery, with the new emphasis on planning for as well as on mitigating effects. Intervenor evidence implies that a potential marine oil spill from the NGP would inevitably have “1989-era” human environmental effects starting point. In reality, if an oil spill was to occur in the future, it would take place within a vastly different human environment than EVOS. Oil spill response efforts are undertaken very differently than the responses twenty-three years ago. What were considered “best practices” in 1989 are not the “best available practices” in 2012 or in the future. Following the lead of the National Oceanic and Atmospheric Agency (NOAA) and the Bureau of Ocean Energy Management (BOEM), two lead regulatory agencies in the U.S., it is clear that that future technologies and response capacities in that jurisdiction will become even more rigorous and refined. Therefore it is reasonable to assume that government oversight and involvement will be even more effective, and that industry progress in this area will continue even as spills become less common.

July 2012 Page C-3

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

The review also examines how adversarial litigation affected the perceived lack of recovery and restoration during EVOS, particularly in the town of Cordova, given the emphasis on the anomalous Cordova experience in certain intervenor evidence. With the exception of Nancy Yaw Davis (1996), social scientists never formally reviewed or evaluated the topic of recovery in the human environment after EVOS until recently (Fall 2009, Picou et al. 2009). Intervenor evidence cite data related to the Exxon Valdez Oil Spill (EVOS) and the Selendang Ayu spill to refute certain findings of the Northern Gateway Environmental and Socio-economic Assessment. For example, the Heiltsuk Tribal Council (A38059), “Heiltsuk Tribal Council Submission of the Potential Effects of an Oil Spill on the Heiltsuk,” points to EVOS as the blueprint for wholesale sociocultural destruction by claiming that, because of EVOS, the sacred connection between the human communities and the oceans was inexorably severed. The Heiltsuk report, like many of the other intervenor reports, claims that not only is a major ENGP marine oil spill inevitable, but also that once a spill occurs, the relationship of Aboriginal people to the environment and each other will be permanently degraded. Intervenor evidence points to the initial post-spill effects of EVOS, primarily the emotionally- charged 1989 effects, the year of the spill, and incorrectly assumes that these effects occurred evenly across the entire spill region, and that that they persist into the present. However, recovery to the human environment is not discussed, even though at least one agency report summarizes spill effects on Alutiiq subsistence use and clearly shows that recovery has occurred (Fall et al. 2001). Fall’s 2001 report summarized over 15 years of research by the Division of Subsistence, Alaska Department of Fish and Game. Through the selective use of source material and “evidence,” intervenor evidence leads to the erroneous conclusion that all Alaska Native people and communities experienced permanent devastating effects across geographically widespread areas of coastal Alaska, that recovery never occurred, and that Aboriginal people in B.C. will experience similar devastating, widespread and irreversible human environmental effects should a marine spill occur. Such evidence does not address post-spill recovery or restoration, and ignores the substantial improvements in emergency oil spill response that have occurred since 1989. Moreover, the evidence ignores changes in the regulatory framework that now include planning and response input from all Alaska Native people in a proposed or real area of effect, as well as active involvement of Aboriginal organizations in marine oil spill planning and response (Pacific States/BC Oil Spill Task Force 2011). As described in Pearson et al. 2012, spill-related effects on the natural environment can be mitigated, recovery of the natural environment does occur, as does recovery of linked social-ecological systems. In neglecting to provide the proper context for understanding the direct, indirect and cumulative effects of EVOS and SAOS, intervenor evidence gives an inaccurate impression that any marine oil spill will inevitably result in unmitigated human environmental devastation. To show that this is not the case, this review discusses aspects related to the EVOS and SAOS showing how, in addition to recovery of the biological environment, recovery has also occurred in the human environment. The review considers the Alutiiq region of Alaska following the EVOS and describes the effectiveness of the response to the SAOS, a response that took

Page C-4 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

immediate steps to mitigate effects on the human environment and foster recovery, both in the short-term and the long-run. One of the primary problems that continues to impair social scientists’ ability to objectively document recovery in the human environment after EVOS was the litigious context of nearly all of the studies and data collection. Now that the legal process has essentially been concluded, litigation-driven effects associated with EVOS are being acknowledged (IAI 2010, Fall et al. 2001), and it is now possible to address the topic of EVOS recovery in a more transparent fashion, as Medred (2010) observes: Privately, there are plenty of people willing to bluntly admit the Sound has recovered. Publicly they are careful to couch their statements to make sure that a reporter might be able to read into them the belief -- reinforced by some in Cordova -- that the spill continues to have major environmental impacts. The fact that there is currently a much different regulatory environment and a more comprehensive and effective preparedness and response climate is critically important when addressing the potential effects of marine spills associated with the NGP. This review explicitly considers and addresses such mitigation.

C.3 Socio-cultural Effects of the EVOS and SAOS This section of the review describes the socio-cultural effects of the EVOS and SAOS based on a comprehensive review of the literature, highlighting the extensive improvements in oil spill preparedness and response that resulted from passage of legislation regulating oil transport in the United States. Both spills occurred in marine waters that are part of the traditional territories of Alaska Native people – the Alutiiq people in Southcentral Alaska and the Unangan (aka Aleut) people in the Aleutian Islands. Both events affected subsistence and other natural resources harvested, used and shared by local Alaska Natives and non-Natives. Both affected local perceptions of food safety and health, and raised concerns about harvesting of natural resources that are important components of the local way of life, and key elements of tradition and cultural identity. Both garnered extensive local and national press coverage. Although the EVOS and SAOS affected the natural environment, the context for each was vastly different. The EVOS resulted in immediate, extensive, high-stakes and lengthy litigation, resulting in an acrimonious relationship among local residents, particularly in the insular community of Cordova. The lack of data sharing about sociocultural issues which were and continue to be pertinent to recovery, particularly by local, permanent and/or seasonal residents who were “clients” of law firms, which controlled access to community residents, further compounded the issue. Multiple explicit and implicit controversies seriously affected the ability of social scientists to conduct unfettered and objective research, with some of this being caused by the social scientists themselves, some by community residents, and to some extent by legal firms and oil industry associates and entities with a vested financial interest in continuing the controversy through costly litigation. The lesson learned for everyone involved with EVOS was

July 2012 Page C-5

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

that the “litigation effects” were not in the long-term best interest of anyone; most of all community residents throughout the EVOS area of effect (see Reynolds 1993, Davis 1996, IAI 2001, and Picou 2009).

C.3.1 Picou, Gill and Ritchie’s Research on Effects of EVOS on Cordova Picou et al. (2001:282) describe how sociologists focus on “the interrelated economic, social and psychological impacts of such events on communities, organizations, families and individuals” when assessing the effects of industrial disasters such as EVOS. Picou, Gill and Ritchie (Picou et al 1992, Gill and Picou 1997; Picou et al 2001) focus their “oil spill effects” research almost exclusively on the effects of EVOS on one community only, Cordova, Alaska, with particular emphasis on “stress” caused by the spill, especially situating their discussion in the context of on-going litigation over punitive damages. On the litigation and its effects, Picou (2009: 88) strategically but unconvincingly positions his argument within a context of “victimization,” a common tactic in anthropology and sociology, but one that loses traction when written from the point of view of a single community, Cordova (2009:88): “The EVOS was the quintessential technological disaster. It was caused by reprehensible human error and resulted in the massive contamination of a pristine natural environment. There was a complete organizational response failure and the ecology continues to be damaged. Furthermore, after twenty years of seemingly mindless litigation, survivors view the final judgment rendered by the Supreme Court as a heinous miscarriage of justice. Ecological and sociological recovery from the EVOS may never occur because of the massive and lingering contamination and the failure of the United States legal system to address the documented damages to the people and communities of Prince William Sound. Research on both the ecological and sociological damages should continue — twenty years after running aground on Bligh Reef, the EVOS continues to manifest serious risks throughout the impact region.” By focusing on a small subset of the entire range of EVOS human environmental effects, Picou, Gil and Ritchie did not address the issue of recovery from EVOS effects on the human environment generally until Picou’s work in 1999 with the Prince William Sound Regional Citizen’s Advisory Council (PWSRCAC 1999). After many years of focusing on stress and “corrosive” issues (the term Gill and Picou 1998 use to characterize the post-spill social effects in which relationships become acrimonious rather than therapeutic), they acknowledged the prospect of recovery (Picou et al. 2009) by noting how communities could set up programs to mitigate some of the negative human environmental effects of large technological disasters like oil spills. The subchapter of Picou et al. (2009:301- 304) entitled “Community Recovery from the EVOS: Educational Intervention as a Mitigation Strategy” includes discussions of a peer listening training program for Cordova residents and a Talking Circle for Eyak residents to help mitigate what they describe as chronic effects of the spill in Cordova.

Page C-6 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

Such intervention strategies are well established in the medical and mental health fields, and are also well known to those working in all aspects of Community Based Participatory Research (CBPR). However, the metrics for measuring outcomes and success are not always explicit, not always clear, and sometimes circular in that the conclusions are given by the premise. Like many Alaska Native groups and communities, the Eyak people had Talking Circles before the EVOS, specifically with respect to education, language loss, and intergenerational conflicts, all problems of concern throughout rural Alaska, including areas inland that have never been involved in any kind of industrial development activity. Indeed, indigenous groups in Alaska, Canada, , New Zealand and Australia use the same approach to “heal,” but perhaps it is the process for them that is most important, even more than any specific products or outcomes that might emerge. In short, the approach undertaken by Picou et al. (2009) was not new. Two salient points can be made regarding this recent inclusion of mitigation efforts in the Picou et al (2009) work: 1. This is consistent with the “upgraded” post Oil Pollution Act of 1990 (OPA 90)3 phenomena of enhanced efforts to plan for and set up mitigation efforts prior to any spills with regard to and/or on behalf of agencies, industries and local governments where oil is produced and shipped in Alaska. As is discussed later in this review, this was done in response to the Selendang Ayu spill. 2. While intervenor evidence on behalf of Aboriginal organizations on the Northern Gateway Project includes effects noted by Picou et al (2009) it does not discuss or call for the implementation of such mitigation measures. Human environmental protection plans in Alaska currently include (but are not limited to) cooperative spill response planning that includes shippers/potentially responsible parties, government agencies, aboriginal organizations, landowners and potentially affected communities/parties. The human environment protection plans are modeled after the Alaska Geographic Response Plans (GRP) map-based spill response plans and activities. The GRP planning effort systematically prioritizes areas for implementation of protection and response measures that are based on community rankings and community inputs, and specifically includes strategies to mitigate potential negative effects during the response. However, it is important to note that the rankings were not established within a context of litigation, acrimony and/or controversy, and are thus potentially more likely to reflect both real and perceived community concerns. Northern Gateway has proposed to develop Geographic Response Plans in cooperation with participating Aboriginal organizations to deal with potential effects from a marine oil spill (See Section C.4 of this review). A component of these plans, similar to those in Alaska, would not only include community priorities for spill response but also mitigation for effects such as changes in traditional harvesting activities, harvesting success, and associated cultural changes.

3 The Oil Pollution Act of 1990 (OPA 90) was formally signed into law in the United States in August 1990. OPA 90 improved the United States’ ability to prevent and respond to oil spills by establishing provisions expanding the US government's ability to respond to oil spills (EPA 2012). OPA 90 also created the National Oil Spill Liability Trust Fund, a fund that provides up to one billion dollars per spill incident, with OPA 90 further requiring that new oil spill contingency planning be conducted in collaboration by both government and industry.

July 2012 Page C-7

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

A recent example of the increased involvement of Federally Recognized Tribes in the decision making process to mitigate cultural effects of marine oil spills is Section 706 of the U.S. Coast Guard Authorization Act. This section requires the USCG to develop a tribal consultation policy regarding oil spill prevention, preparedness, response, and damage assessment, and further states that, if an oil spill is likely “to have a significant impact” on natural and cultural resources owned or used by a federally recognized tribe, the USCG must include tribal representatives in the incident command system; share information about the spill; and involve the tribe in response decisions (Pacific States/British Columbia Oil Spill Task Force 2011). The NOAA Office of Oil Spill Response and Recovery is implementing similar policies to address issues related to offshore oil development, in the Bering, Chukchi and Beaufort Seas in Alaska. Additionally, ongoing oil spill response planning for transboundary areas in Washington, Alaska and B.C. provides guidance for involving Aboriginal organizations in spill planning and response.

C.3.2 EVOS Studies on Human Environment Effects Three major socio-economic research initiatives were conducted following the EVOS: • the Oiled Mayors Project (Palinkas et al.. 1993; IAI 1990a, 1990b, 1990c, 1990d; Russell et al.. 1996); • the MMS-sponsored social indicators study (Human Relations Area Files 1992-1995) and • the National Science Foundation sponsored Cordova Community Study (Picou and Gill 1996). The three initiatives addressed very different aspects of the spill for different purposes, including litigation, and employed very different research designs and methods. The Oiled Mayors Project, funded by the State of Alaska Department of Community and Regional Affairs, was a one-year intensive study geared toward understanding the types and range of social, economic and psychological effects of the spill and cleanup on 22 Alaskan communities (see Figure C-1 for the geographic distribution of the 22 communities).

Page C-8 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

Source: (adapted from Looking Both Ways: Heritage and Identity of the Alutiiq People of Southern Alaska). http:// www.mnh.si.edu/lookingbothways

Figure C-1 The Alutiiq Region Alutiiq Region Villages One aspect of the Oiled Mayors Project that received subsequent amplification by social scientists and that has been the focus of much of the Aboriginal organizations’ objections to NGP, concerns the psychological effects of the spill (Palinkas et al. 1993). Consider, for example, the following comment from that publication: A progressive "dose-response" relationship was found between exposure to the oil spill and the subsequent cleanup efforts and the following variables: reported declines in traditional social relations with family members, friends, neighbors and coworkers; a decline in subsistence production and distribution activities; perceived increases in the amount of and problems associated with drinking, drug abuse, and domestic violence; a decline in perceived health status and an increase in the number of medical conditions verified by a physician; and increased post-spill rates of generalized anxiety disorder, post-traumatic stress disorder, and depression. Alaskan Natives, women, and 18-44 year olds in the high- and low-exposed groups were particularly at risk for the three psychiatric disorders following the oil spill. The results suggest that the oil spill's impact on the psychosocial environment was as significant as its impact on the physical environment.

July 2012 Page C-9

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

This was a single-point-in-time study that was not intended to track long-term trends or to deal with recovery issues related to EVOS. Davis (1996:246) was critical of the lack of interest in the topic of recovery studies when she summarized the results of the “Oiled Mayors” and MMS studies, noting “both indicate that greater disruption occurred in the smaller, Native, communities. That finding may have been the result of the timing of the research and the general orientation that anthropologists tend to have of concern for smaller Native communities. It also may be an artifact of questions that seem designed to understand kinds and degrees of impact. That was, after all, their charge.” The Cordova Community Study (Picou and Gill, 1996) was intended neither to measure broad patterns of effects nor to gauge recovery. Rather it involved attempts to measure long-term psychological effects of the spill in terms of community response. The sociological methodology included telephone surveys geared toward assessing patterns of “community-environment disruptions” that certain natural-resource dependent communities’ experience, along with the long-term collective stress that these effects engender (Picou et al.. 1992; Picou and Martin, 2007). However, there is a clear confusion here between cause and effect, and an arbitrary if not ambiguous time line and chronological frame of reference that do not immediately inform the conclusions drawn. All people through the EVOS area of effect were clearly and differentially affected to be sure, but many of the social, mental health and behavioral issues attributed to the spill have their origins in a “post colonial context,” which occurred long before the spill. While it is acknowledged that in the short run, and perhaps even in the long run, the referenced social problems are prevalent, two contextual points must be kept in mind: • To argue that these problems were directly caused by the spill is difficult to support or defend with available data; • It is difficult to gauge these effects given that these same problems occur in areas well outside the area of the EVOS. It is just as easy to argue that many of these problems reflect the combined, cumulative effects of colonization, acculturation, and lack of economic opportunities, to name but a few important effects (Loring and Gerlach 2010). A decline in subsistence activities by Alaskan native people, oil spills notwithstanding, started long ago with the nutrition transition, and is exacerbated now by the rising price of fuel that makes it difficult for people across Alaska to access country foods. Similar trends have been noted in Canada. In addition, there are intergenerational issues with some younger people not interested in traditional cultural and food procurement activities. In Alaska, it has been argued that post-traumatic stress disorder exacerbated all social and cultural problems with the Bureau of Education, Collier’s program of Assimilation, beginning with IRA in 1934. As a result of this initiative, the current generation of Elders, in their youth, was sent away to school, in many cases leaving the villages for years, and returning in many cases without much local knowledge about how to cope with life in the villages or how to live effectively with the land and seascape. It also could be argued that the cause and effect linkage is seductively cogent, but empirically difficult to substantiate objectively. These aspects are well known and acknowledged by many Alaska Native leaders and Elders today, and are similar to the residential school issues in Canada.

Page C-10 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

Picou and Gill (1996) focus on the “renewable resource community” concept in which they reported that the socioeconomic effect of the spill on a community was positively associated with how strongly that community depended on renewable natural resource harvests. This is a reasonable argument, and one that has been argued by others. Picou and Gill (1996) documented higher levels of stress in the community of Cordova, which had a high dependence on renewable resource harvests than in the more socioeconomically diverse nearby community of Valdez or the “control” community of Petersburg (in Southeast Alaska), but again, the inherent ambiguity in cause and effect relationship confuses rather than clarifies Subsequent research (Picou and Martin, 2007), conducted 17 years after the spill, claimed that members of affected communities were still exhibiting signs of chronic psychological and community disruption. However, this also could be the result of cumulative effects of long-term social and cultural change, versus only effects arising from the EVOS. There are many communities from the Kenai Peninsula to the Aleutians that exhibit the similar types of change and community characteristics, but have never been negatively affected by industrial development. In this context, EVOS is probably best viewed as a contributing factor in stress rather than the singular causal factor. Several authors (Fall et al. 2001; IAI 2001) have noted that stress also resulted from the extended and adversarial litigation process being used to obtain punitive damage compensation from Exxon, litigation that is acknowledged to have been stressful and damaging to the communities around Prince William Sound. The Impact Assessment Incorporated (2001) four volume report summarized the results of the three major EVOS research initiatives and includes a factor-by-factor analysis including the cultural and economic effects. This comprehensive series of reports considers the direct socio- economic effects caused by the spill itself, as well as the indirect effects resulting from the cleanup and subsequent litigation. These reports are extensive, including volumes of data regarding all aspects of cultural, economic and social effects of EVOS through time and throughout Alaska, with a particular emphasis on the role of oil development that could never be captured by sociological observations from a single community.

C.3.3 Recovery in the Human Environment – the EVOS Experience Davis (1996:233) conducted a quite challenging anthropological study of the EVOS effects in 1989 for the North Pacific Rim (now known as Chugachmuit). In contrast to Picou et al., Davis had extensive prior experience in the communities researching potential effects of oil and gas development, and was intimately familiar with the communities’ response to and recovery from the Great Alaska Earthquake of 1964 which destroyed Chenega village and resulted in extensive demographic and cultural adjustments. What is significant in her published article regarding EVOS (1996), an article that summarizes observations based on decades of research in the region, is her reflections on the endurance, adaptability and resilience of Alutiiq communities in the face of disasters such as , and volcanoes that are a part of life in the region. Davis (1996: 252) further notes that the class action lawsuits filed on behalf of fishermen and Native populations constrained social science research associated with the recovery of

July 2012 Page C-11

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

communities from the spill, and, again, addresses the confusion in understanding and disentangling the effects of both the spill and litigation: . . . if the Alutiiq communities survived the 1912 Katmai eruption and the 1964 earthquake, it may be safe to assume that they will also survive the consequences of the oil spill. If the people of Valdez have managed to rebound from a gold rush, a fire, an earthquake, and a , we must assume they will also survive Exxon Valdez, perhaps exceptionally well. But social scientists who want to analyse the human responses to this oil spill are severely handicapped because there are virtually no funds for understanding that process. Moreover, pending litigation has put a damper on scientific research of all kinds. The few investigators who might be in a position to offer informed assessments are discouraged from offering their views by the threat of legal penalties. Davis also clearly exposes the context of the “Day the Water Died” speech that Gill and others cite in intervenor evidence to the JRP (and elsewhere – see Gill and Picou 1997) , and reveals the inner workings of the legal process in promoting a perception of human environmental damage without the prospect of recovery (1996:255): Native adaptation strategies of creative accommodation to changes provide a challenge for attorneys, who seek evidence of social, mental and economic damage linked to Exxon Valdez in order to win funds for their clients and fees for themselves. A classic example of coaching to say the politically correct thing in public occurred during the summer of 1989. A “chief” who was billed as a high subsistence user of a small village supposedly inundated with oil, was scheduled to read a speech at a meeting of international “oiled mayors.” An attorney wrote a rather powerful emotional narrative but the chief did not read it. Nor did he attend the meeting. The oil had not damaged his resources: he was a relatively wealthy man and no longer participated in the rigour of subsistence harvesting and processing: he bought his groceries in a store. Later that fall, his village had a hard time processing all the free fish that was shipped at great expense. The freezers were already full of store-bought goods. Furthermore, significant amounts of oil never reached this community. But social services and attorneys did. The chief was too kind to hurt their feelings and, for a while, simply accepted what others insisted on bringing – including a speech he is credited with but did not write. In her conclusions about recovery of the human environment after EVOS, Davis (1996:264) noted that the three major disasters, the1912 Katmai eruption, the 1964 earthquake, and the1989 oil spill that hit the region were the three largest of their kind. Each one led to new understandings of the natural and human environment. While some people typically think of disasters as “destroyers” of cultures, Davis notes how and under what conditions they can provide building blocks for new social organization and new directions through community based creativity, innovation and effective responses to change. She also notes the resilience of the local communities:

Page C-12 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

I suggest that small communities have greater abilities for recovery than are usually recognized. They have kinship, family, history, identity, past experiences, local leadership, all combined and activated by a major event. Resolution of local pre-disaster conflicts may need to occur as part of recovery. External involvement by attorneys, social services, and even inappropriate research may inhibit recovery processes and prolong or sharpen awareness of the original pain, extending in time the negative impacts. Fall et al. (2001:2) produced a document for the MMS (and also printed as ADFG Division of Subsistence Technical Report # 264) which described how the communities affected by EVOS actively adapted to the spill in ways that protected family well-being and preserved traditional cultural elements unique to the local way of life. Most of their information was collected directly from residents throughout the spill area, including through systematic face-to-face household surveys in the communities and through ethnographic research over a period of 10 years after the spill. In contrast to descriptions of oil spill effects included in intervenor evidence to the Northern Gateway JRP, including, unfortunately, terms such as “genocide”, Fall et al. (2001:291-305) summarize the effects as temporary destabilizing, but not culturally cataclysmic or catastrophic in the long-run. Their analysis is based on an objective and clear evaluation of over ten years of data regarding the effects on the local food supply, food safety, support networks, replacement of lost harvests, wage and income effects, economic and cultural revitalization, demographic trends, and dispossession of Native lands and the survival of Alutiiq communities. Fall and colleagues state (2001:2): Most of the information derives directly from residents of the spill area provided through several systematic face-to-face household surveys in Pacific Gulf communities and through ethnographic research . . . the findings strongly indicate that in most communities, families actively adapted to the industrial disaster in ways that protected the well-being of family members, and that preserved traditional cultural elements in the community’s way of life. Most Alutiiq villages appear to have endured the disaster through the hard choices and work of extended families and tribal governments. This outcome is a testament to the durability of traditional ways of living that support contemporary Alaska Native communities. The Fall et al. (2009) report is among the few published government-funded EVOS reports that systematically consider litigation and community recovery together – two topics that were previously addressed only minimally by prior EVOS human environment studies. In Fall et al (2009), the authors summarize 15 years of ADF&G Subsistence Division research related to spill-related changes in local use of fish and game – by far the most extensive suite of EVOS human environment studies. The intervenor evidence that Gill and Ritchie provide lacks the broader understanding of post- spill human environmental recovery because they focus almost exclusively on negative effects. Additionally, Gill and Ritchie’s assessment of the Selendang Ayu (2006) – in which they found

July 2012 Page C-13

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

scant evidence of their model’s predicted devastating sociocultural effects related to this spill – further indicates that their framework, methods and methodology does not account for the reality of improved oil spill preparedness and response that can mitigate negative effects on the human environment.

C.3.4 The Selendang Ayu - a Post-EVOS Example of Recovery in the Human Environment In Alaska (and other coastal areas of the United States), the Bureau of Ocean Energy Management, Regulation and Enforcement (BOEMRE) is responsible for administering oil and gas development on the outer continental shelf per the Outer Continental Shelf Lands Act of 1953 (OCSLA). The BOEMRE is guided by the National Environmental Policy Act of 1969 (NEPA) which helps inform U.S. environmental policy decisions using research on the human (and natural) environments. Both NEPA and OCSLA authorize BOEMRE to conduct and sponsor studies of coastal and marine environments (including human environment) potentially affected by oil and gas industry activities occurring on the OCS. BOEMRE Alaska OCS Region administers the Alaska Environmental Studies Program (ESP) to “define information needs and implement studies to assist in predicting, projecting, assessing, and managing the potential effects of oil and natural gas development on the human, marine, and coastal environments of the OCS and [adjacent] coastal areas” (MMS 2002:1). Information from OCS studies is used for decision making and planning purposes, including those associated with agency Environmental Assessment (EA) and Environmental Impact Statement (EIS) documentation (IAI 2010:2). EVOS led to the passage of the Oil Pollution Act of 1990 (OPA 90), that was formally signed into law in the United States in August 1990. OPA 90 improved the United States’ ability to prevent and respond to oil spills by establishing provisions expanding the US government's ability to respond to oil spills (EPA 2012). OPA 90 also creates the national Oil Spill Liability Trust Fund, a fund that provides up to one billion dollars per spill incident. OPA 90 also required that new oil spill contingency planning be conducted in collaboration by both government and industry. Area Committees, composed of federal, state, and local government officials, were required to develop Area Contingency Plans. Owners or operators of vessels and certain facilities also were required to prepare response plans. In combination, this new regulatory framework reflects a new philosophy, and a “new way of doing business,” . By the time the Selendang Ayu went aground in a storm off of Unalaska Island in 2005, the response regime had changed dramatically from the one in place during EVOS in 1989, in part because of upgraded spill response preparedness and planning. Public and private response efforts for spill events were in place, enabling an improved response and the ability to openly address recovery. Many of the Selendang Ayu responders including federal (US Coast Guard, NOAA, USFWS), state (Alaska Department of Environmental Conservation, DNR), and local government (City of Unalaska) officials; various tribes (Qawalangan Tribe), and the ship owner’s representatives had access to Area Contingency Plans. Most either had EVOS experience, and/or in-depth training in Incident Command System (ICS) and Shoreline Cleanup

Page C-14 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

Assessment Team (SCAT) spill response. A number of individuals in the Unified Command structure had been involved in the cleanup associated with a previous shipwreck (the M/V Kuroshima) on Unalaska Island in 1997, giving the Selendang Ayu responders a relatively high comfort level regardless of affiliation. When the Selendang Ayu spill occurred in December 2004, the National Oil and Hazardous Substances Pollution Contingency Plan (the National Contingency Plan or NCP), the federal government's blueprint for responding to oil spills, had been developed to improve national response capability and promote overall coordination among the hierarchy of responders and contingency plans. The NCP included many of the Selendang Ayu responders. The Selendang Ayu spill response involved a highly integrated Unified Command, experienced response personnel trained in ICS, and a community of responders aware of the importance of natural resource and human environmental sensitivities. The responders employed many strategies to help mitigate effects on the human environment. The shipping company’s lead representative (IMC’s Howard Hile) helped the Unalaska community conduct a Subsistence Forum in early 2005 in which Father Michael Oleksa delivered the keynote address “Communicating Across Cultures: Subsistence and Science” at the opening reception. Other activities included a “Subsistence Fair” with a kid’s corner, arts and crafts displays, touch tank aquarium, video presentations, information on the spill response, and many other exhibits. This was followed by a community feast that included locally-harvested, culinary and culturally important delicacies prepared using traditional methods and recipes. Interpretive beach walks and tours were conducted with a variety of experts in the afternoon, covering topics such as tide-pooling, shoreline cleanup and assessment (SCAT), invertebrate ecology, medicinal plants, subsistence gathering, and hatcheries. Closing remarks and discussion were facilitated by Father Oleksa (ADEC 2005). This event had the desired effect of bringing responders and community members into a respectful dialogue that continued throughout the response (Morris 2005). The ensuing dialogue was similar to the “talking circle” intervention strategy proposed by Picou and Gill. Another example of how to prevent impacts on the human environment was the Selendang Ayu’s daily “Incident Action Plan” (IAP). The IAP is one of the most important documents produced by the ICS because it provided the operational plans necessary to execute an oil spill response through the clear and explicit incorporation of key stakeholder concerns. The second priority in the IAP, after protecting the health and safety of the public and the responders, was “Protect sensitive areas to minimize impact to the environment, cultural, subsistence, and economic resources and property (Morris 2005).” While the Selendang Ayu spill was not on the scale of the EVOS event; lessons learned from the EVOS have been incorporated into the oil spill response in the regulatory and preparedness realms. The success of the Selendang Ayu response effort demonstrates that it is not reasonable to directly compare the effects of the 1989 EVOS event on the human environment with events that take place under current response and preparedness plans – much less those that might occur over the course of the next 30 years, as those plans are improved and refined.

July 2012 Page C-15

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

C.4 Northern Gateway Project Geographic Response Plans Northern Gateway has committed to the development of Geographic Response Plans with opportunity for community participation. This is clearly in keeping with the types of mitigation measures that have been and continue to be refined and developed in Alaska. The socio-economic assessment for the Northern Gateway Project in Section 9.5 of Volume 8C acknowledges the range of social and economic effects that could result from an accidental spill or release. It focuses on three types of potential effects: • Adverse quantifiable effects experienced by individuals, Aboriginal communities or companies • Adverse effects on the social well-being of residents of nearby renewable resource communities, where community members are particularly vulnerable to oil spills. • Adverse effects that clean-up and remediation activities can have on local communities and infrastructure. As part of its commitment to oil spill response planning, Northern Gateway will include in its response plans measures to address socio-economic effects in cooperation with government agencies, participating Aboriginal organizations and non-aboriginal communities and businesses. These will include measures to prioritize emergency response and remediation activities, as well as to facilitate replacement of traditional foods, and compensation for fishing and harvesting losses. These would leverage other program commitments such as the supplementation of environmental and traditional use information through collection of harvesting data, community participation in preparation of environmental sensitivity atlases, and environmental effects monitoring programs prior to commencement of operations.

C.5 Conclusions If intervenor experts base their predictions of potential spill effects on the human environment on an EVOS type scenario, they are using obsolete information. It is more precise and appropriate to use the Selendang Ayu spill response experience, and to reference the preparedness regime that exists in other Alaskan GRS/ICS savvy communities as the current state of the art for oil spill response, mitigation and recovery in the human environment. As shown in the SAOS, approaches such as ICS training, cooperative and transparent SCAT processes, engagement of Aboriginal people and non-aboriginal community members in response planning, integration of cultural and cultural resource sensitivity into response plans, food safety testing, and dissemination of culturally appropriate information on such testing can mitigate adverse effects on the human environment. Given that Northern Gateway has committed to develop similar types of approaches and plan in cooperation with government agencies and participating Aboriginal organizations and non-Aboriginal community members, adverse effects on the human environment can be addressed and recovery of the human environment would be facilitated before, during and after a potential spill event.

Page C-16 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

C.6 References ADEC. 2005. Alaska Department of Environmental Conservation Division of Spill Prevention and Response Prevention and Emergency Response Program. Situation Report #73, M/V Selendang Ayu Incident, April 14, 2005. Chorney, N.E., A.B. McCrodan, and M.E. Austin. 2011. Northern Gateway Pipeline Project: Vessel Transit Noise, Marine Acoustics Modelling Study, 2011. Version 2.0. Technical report prepared for Stantec Consulting Ltd. for Northern Gateway Pipeline Project by JASCO Applied Sciences, January 2011.Fischer, R.W. and N.A. Brown. 2005. Factors affecting the underwater noise of commercial vessels operating in environmentally sensitive areas. OCEANS 2005. Proceedings of MTS/IEEE 3: 1982–1988. Davis, N.Y. 1996. The Exxon Valdez Oil Spill, Alaska. In: The Long road to Recovery : community responses to industrial disaster. Edited by James K. Mitchell. United Nations University Press. Fall, J.A. 1999. Subsistence. Restoration Notebook Series. Exxon Valdez Oil Spill Trustee Council. Anchorage Fall, J.A., R. Miraglia, W. Simeone, C. Utermohle, and R.J. Wolfe. 2001. Long-Term Consequences of the Exxon Valdez Oil Spill for Coastal Communities of Southcentral Alaska. James A. Fall (ed.).Technical Report No. 163. OCS Study MMS 2001-032. Final Report for: Sociocultural Consequences of Alaska Outer Continental Shelf Activities: Data Analysis and Integration. Anchorage. Fall, J.A., R.J. Walker, R.T. Stanek, W.E. Simeone, L. Hutchinson-Scarborough, P.Coiley-Kenner, L. Williams, B. Davis, T. Krieg, B. Easley, and D. Koster. 2006. Update of the Status of Subsistence Uses in Exxon Valdez Oil Spill Area Communities, 2003. James A. Fall (ed.). Technical Paper No. 312. Alaska Department of Fish and Game, Division of Subsistence. Juneau. Gill D. and L. Ritchie. 2006. The Selendang Ayu oil spill: a study of the renewable resource community of Dutch Harbor/Unalaska. Quick response research report #181, University of Colorado, Boulder. Natural Hazards Research and Applications Information Center. Gill, D. and S. Picou. 1997. The day the water died: cultural impacts of the Exxon Valdez oil spill. In The Exxon Valdez disaster: readings on a modern social problem, eds. J.S. Picou, D.A. Gill, and M.J. Cohen, 167 1998. Technological Disaster and Chronic Community Stress. Society and Natural Resources 11, 795- 815.-187. Dubuque, IA: Kendall/Hunt Publishing Co. Human Relations Area Files, Inc. 1995. Social Indicators Study of Alaskan Coastal Villages, Volume VI. Analysis of the Exxon Valdez Spill Area, 1988-1992. Prepared by J. Jorgensen, Principal Investigator. U.S. Department of the Interior, Minerals Management Service, Alaska OCS Region Social and Economic Studies Technical Report No. 157. (MMS 94-0064)

July 2012 Page C-17

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

1994a. Social Indicators Study of Alaskan Coastal Villages, Volume III. Analysis. J. Prepared by Jorgensen, Principal Investigator. U.S. Department of the Interior, Minerals Management Service, Alaska OCS Region Social and Economic Studies Technical Report No. 154. (MMS 93-0070) (Contract No. 14-12-0001-30300) 1994b. Social Indicators Study of Alaskan Coastal Villages, Volume V. Research Methodology for the Exxon Valdez Spill Area, 1988-1992. Prepared by J. Jorgensen, Principal Investigator, and S. McNabb, Senior Investigator. U.S. Department of the Interior, Minerals Management Service, Alaska OCS Region Social and Economic Studies Technical Report No. 156. (MMS 93-0071) (Contract No. 14-31-0001-30300). 1993a. Social Indicators Study of Alaskan Coastal Villages, Volume II. Research Methodology: Design, Sampling, Reliability, and Validity. Prepared by J. Jorgensen, Principal Investigator. U.S. Department of the Interior, Minerals Management Service, Alaska. OCS Region Social and Economic Studies Technical Report No. 153. (MMS 93-0035). 1993b. Social Indicators Study of Alaskan Coastal Villages, Volume IV. Postspill Key Informant Summaries. Schedule C Communities, Part 1 (Cordova, Tatitlek, Valdez) and Part 2 (Kenai, Tyonek, Seldovia, Kodiak City, Karluk, Old Harbor, Chignik). Prepared by J. Endter-Wada, J. Hofmeister, R. Mason, S. McNabb, E. Morrison, S. Reynolds, E. Robbins, L. Robbins, and C. Rooks. J. Jorgensen, Principal Investigator. U.S. Department of the Interior, Minerals Management Service, Alaska OCS Region Social and Economic Studies Technical Report No. 155. (MMS 92-0052) (Contract No. 14-12-0001-30300). 1992a. Social Indicators Study of Alaskan Coastal Villages, Volume I. Key Informant Summaries, Volume 1: Schedule A Regions, (North Slope, NANA, Calista, Aleutian- Pribilof). Prepared by T. Brelsford, A. Fienup-Riordan, J. Jorgensen, S. McNabb, P. Petrivelli, L. Robbins, and M. Galginaitis. J. Jorgensen, Principal Investigator. U.S.Department of the Interior, Minerals Management Service, Alaska OCS Region Social and Economic Studies Technical Report No. 151. (MMS 92-0031) (Contract No. 14-12-0001-30300). 1992b. Social Indicators Study of Alaskan Coastal Villages, Volume I. Key Informant Summaries, Volume 2: Schedule B Regions, (Bristol Bay, Kodiak, Bering Straits). Prepared by J. Endter- Wada, J. Hofmeister, R. Mason, Steven McNabb, J. Mulcahy, and L. Robbins. J. Jorgensen, Principal Investigator. U.S. Department of the Interior, Minerals Management Service, Alaska OCS Region Social and Economic Studies Technical Report No. 152. (MMS 92-0032) (Contract No. 14-12-0001-30300). Impact Assessment Inc. 2001. Exxon Valdez Oil Spill, Cleanup, and Litigation: A Collection of Social Impacts Information and Analysis. Final Report. MMS OCS 2001-058. Prepared for the U.S. Department of the Interior, Minerals Management Service, Alaska OCS Region. Anchorage. 2010. Social and Economic Assessment of Major Oil Spill Litigation Settlement Final Baseline Report. Prepared for U.S. Department of the Interior, Bureau of Ocean Energy, Management, Regulation and Enforcement. La Jolla, California.

Page C-18 July 2012

Attachment 8 to Northern Gateway Reply Evidence

Recovery of the Biophysical and Human Environments from Oil Spills Enbridge Northern Gateway Project Appendix C: Recovery of the Human Environment from Oil Spills: Alaskan Case Studies

Jordan, T., K. Hollingshead, M.T. Keevin and G.L. Hempen. 2005. Miami Harbor Deepening - Phase II Dodge-Lummus Island Turning Basin Project and Fisherman’s Channel: Results of Mitigation and Monitoring Measures - June 25 through August 12, 2005 to Protect Dolphins and Manatees During Underwater Blasting. US Army Corps of Engineers. Miami, FL. MacGillivray, A. 2010. Gateway Environmental Management Tanker and Escort Tug Source Level Measurement Study, Valdez Alaska, 2010. Technical report prepared for Stantec by JASCO Applied Sciences, December 2010. Morris, R. 2005. Overview of M/V SELENDANG AYU Response. By Captain Ron Morris, USCG Retired. Presentation to the Aleutian Life Forum, Unalaska, Alaska, 17 August 2005. Pacific States/British Columbia Oil Spill Task Force. 2011. The Stakeholder Workgroup Review of Planning and Response Capabilities for a Marine Oil Spill on the U.S./Canadian Transboundary Areas of the Pacific Coast Project Report. Palinkas, L., M. Downs, J. Petterson, J Russell. 1993. Social, Cultural, and Psychological Impacts of the Exxon Valdez Oil Spill. Human Organization, Volume 52, Number 1, Spring 1993. Picou, J. 2009. When the Solution Becomes the Problem: The Impacts of Adversarial Litigation on Survivors of the Exxon Valdez Oil Spill. University of St. Thomas Law Journal Volume 7, Issue 1 2009 Article 5. Picou, J., Formichella B. and C. Arata. 2009. Community Impacts of the Exxon Valdez Oil Spill: A Synthesis and Elaboration of Social Science Research. In Braund, S and J. Kruse (Eds.), Synthesis: Three decades of social science research on socioeconomic effects related to offshore petroleum development in coastal Alaska (MMS OCS Study Number 2009-006, pp. 279-307). Anchorage, AK: Minerals Management Service, Alaska, OCS Region.) Picou, J.and D. Gill. 1996. The Exxon Valdez Oil Spill and Chronic Psychological Stress. American Fisheries Society Symposium 18, 879-893. Picou, J., D. Gill, C. Dyer and E. Curry. 1992. “Stress and Disruption in an Alaskan Fishing Community: Initial and Continuing Impacts of the Exxon Valdez Oil Spill.” Industrial Crisis Quarterly 6:235- 57. Picou, J. and Martin, C. 2007. Long-term impacts of the Exxon Valdez oil spill: Patterns of social disruption and psychological stress seventeen years after the disaster. Report prepared for the National Science Foundation. Mobile: University of South Alabama. Reynolds, S. 1993. Effects of the 1989 Exxon Valdez oil spill on Cordova, Alaska. Social indicators study of Alaskan coastal villages: IV. Postspill key informant summaries; schedule C communities, Part 1 (Cordova, Tatitlek, Valdez). Prepared for Minerals Management Service, Alaska OCS Environmental Studies Program, Technical Report 155, OCS Study MMS 92-0052. New Haven, CT: Human Relations Area Files

July 2012 Page C-19

Attachment 8 to Northern Gateway Reply Evidence