Fisheries and Oceans Canada (DFO) Annex 3

PDF page

Federal Government, Technical Report: Properties, Composition and Marine Spill 2 Behaviour, Fate and Transport of Two Diluted Bitumen Products from the Canadian Oil Sands; Cat. No. En84–96/2013E–PDF; Government of Canada: Ottawa, ON, Canada, pp. 1– 85, 2013. ISBN 978-1-100

Félix, F., and Van Waerebeek, K. 2005. Whale mortality from ship strikes in Ecuador and 90 west Africa. Latin Amer. J. Aquat. Mamm. 4(1): 55-60.

Fingas, M. 2013. Modeling oil and petroleum evaporation. Journal of Petroleum Science 96 Research, 2(3), 104-115.

Fisheries and Oceans Canada (DFO). 2009. Recovery potential assessment for basking 108 shark in Canadian Pacific waters. Canadian Science Advisory Secretariat. Science Advisory Report 2009/046. 10 pp.

Fisheries and Oceans Canada (DFO). 2009. Recovery potential assessment for Transient 118 Killer Whales. Canadian Science Advisory Secretariat. Science Advisory Report 2009/039. 11 pp.

Fisheries and Oceans Canada (DFO) 2014. Recovery potential assessment for Offshore 129 Killer Whales off the Pacific coast of Canada. Canadian Science Advisory Secretariat. Science Advisory Report 2014/047. 17 pp.

Fisheries and Oceans Canada (DFO). 2015a. Sufficiency review of the information on 146 effects of underwater noise and the potential for ship strikes from Marine Shipping on Marine Mammals in the Facilities Application for the Trans Mountain Expansion Project. DFO Can. Sci. Advis. Sec. Sci. Resp. 2015/007.

Fisheries and Oceans Canada (DFO). 2015b. Technical review of predicted effects and 159 proposed mitigation of underwater noise and potential vessel strikes on marine mammals, from the December 2013 Facilities Application and supplemental information for the Trans Mountain Expansion Project. DFO Can. Sci. Advis. Sec. Sci. Resp. 2015/022.

Fisheries and Oceans Canada (DFO). 2016. Integrated Biological Status of Southern British 174 Columbia Chinook Salmon (Oncorhynchus tshawytscha) Under the Wild Salmon Policy. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2016/042.

Fisheries and Oceans Canada (DFO). 2018. Recovery Strategy for the Northern and 189 Southern Resident Killer Whales (Orcinus orca) in Canada. Species at Risk Act Recovery Strategy Series, Fisheries & Oceans Canada, Ottawa, x + 84 pp.

Fisheries and Oceans Canada (DFO). 2018. Science information to support consultations 283 on BC Chinook Salmon fishery management measures in 2018. DFO Can. Sci. Advis. Sec. Sci. Resp. 2018/035.

001 Fitzpatrick, F., Boufadel, M., Johnson, R., Lee, K., Graan, T., Bejarano, A., Zhu, Z,. 309 Waterman, D., Capone, D., Hayter, E., Hamilton, S., Dekker, T., Garcia, H., and Hassan, J. 2015. Oil-particle interactions and submergence from crude oil spills in marine and freshwater environments—Review of the science and future science needs: U.S. Geological Survey Open-File Report 2015–1076, 33 p., http://dx.doi.org/10.3133/ofr20151076. ISSN 2331-1258.

Ford JKB, and Ellis G.M. 2006. Selective foraging by fish-eating killer whales Orcinus orca 353 in British Columbia. Marine Ecology Progress Series 316: 185-199.

Ford, J.K.B., Wright, B.M., Ellis, G.M., and Candy, J.R. 2010. Chinook salmon predation by 368 resident killer whales: seasonal and regional selectivity, stock identity of prey, and consumption rates. DFO Canadian Science Advisory Secretariat Research Document 2009/101. iv + 43 pp.

002 Government Gouvernement l♦I of Canada du Canada

Federal Government Technical Report

Properties, Composition and Marine Spill Behaviour, Fate and Transport of Two Diluted Bitumen Products from the Canadian Oil Sands

Environment Canada Emergencies Science and Technology

Fisheries and Oceans Canada Centre for Offshore Oil, Gas and EnergyResearch

Natural Resources Canada CanmetENERGY

November 30, 2013

Canada

003 ISBN 978-1-100-23004-7 Cat. No.: En84-96/2013E-PDF

Information contained in this publication or product may be reproduced, in part or in whole, and by any means, for personal or public non-commercial purposes, without charge or further permission, unless otherwise specified.

You are asked to:

• Exercise due diligence in ensuring the accuracy of the materials reproduced; • Indicate both the complete title of the materials reproduced, as well as the author organization; and • Indicate that the reproduction is a copy of an officialwork that is published by the Government of Canada and that the reproduction has not been produced in affiliation with or with the endorsement of the Government of Canada.

Commercial reproduction and distribution is prohibited except with written permission from the author. For more information, please contact Environment Canada's Inquiry Centre at 1-800-668-6767 (in Canada only) or 819-997-2800 or email to [email protected].

Cover Photos: © Author, Environment Canada

Aussi disponible en frans;ais

004 Table of Contents Executive Summary ...... 5 1.0 Introduction ...... 7 2.0 Literature Review of Diluted Bitumen and Synthetic Bitumen Spills, Behaviour and Use of Chemical Countermeasures ...... 11 2 .1 Properties and Composition of DilutedBitumen Products from the Canadian Oil Sands ... 11 2.1.1 DilutedBitumen Products ...... 11 2.2 Fate andBehaviour ofDiluted Bitumen Products and Related Heavy Oils in the Marine Environment ...... 15 2.2.1 Fate and Behaviour of Diluted Bitumen Products in Spills ...... 15 2.2.2 Fate and Behaviour of Orimulsion in Spills ...... 17 2.2.3 Fate and Behaviour of Heavy Oils in Spills ...... 18 2.3 Use of Dispersants on Potential Spills of DilutedBitumen and Related Heavy Oils ...... 21 2.3.1 Use ofDispersants on aDiluted Bitumen Spill ...... 21 2.3.2 Use ofDispersants on Orimulsion ...... 21 2.3.3 Use ofDispersants on Heavy Fuel Oils ...... 21 3.0 Physical Characteristics and Chemical Composition of Diluted Bitumen Products ...... 25 3.1 Samples ...... 25 3.2 Methods ...... 25 3.2.1 Evaporative Weathering ...... 25 3.3 Characterization of Diluted Bitumen Products ...... 26 3.4 Physical Properties of Diluted Bitumens ...... 27 3.4.1 Properties ofDilutedBitumens ...... 31 3.4.2 Pan Evaporation of Diluted Bitumens ...... 33 3.4.3 Simulated Distillation ofDiluted Bitumens ...... 34 3.4.4 Tendency to form Water-in-Oil Emulsions ...... 36 3.5 Chemical Composition of Diluted Bitumens ...... 36 4.0 Oil Weathering and Buoyancy in Marine Conditions ...... 42 4.1 Introduction...... 42 4.2 Materials and Methods ...... 43 4.2.1 Oil Types ...... 43 4.2.2 Evaporation ...... 43 1

005 4.2.3 Photo-oxidation...... 44 4.2.4 Oil-Water Mixtures ...... 44 4.2.5 Oil-Water-Sediment Mixtures ...... 45 4.3 Results...... 46 4.3.1 Evaporative Weathering ...... 46 4.3.2 Photo-oxidative Weathering ...... 47 4.3.3 Mixingwith Salt Water...... 48 4.3.4 Mixingwith Salt Water and Sediment...... 50 4.4 Comparisons with Literature and Historical Spills ...... 54 5.0 Wave Tank Studies: The Fate and Behaviour of Selected Diluted Bitumen Products in Natural Sea States and Environmental Conditions ...... 56 5.1 Introduction...... 56 5.2 Materials and Methods...... 56 5.2.1 Wave Tank Studies...... 56 5.2.2 Oil, Dispersant and Mineral Fines Application ...... 56 5.2.3 Wave Tank In-situ Measuring Devices ...... 57 5.2.4 Lab Analysis of Seawater Samples ...... 57 5.3 Results and Discussion...... 58 5.3.1 Synthetically Weathered Dilbit Products ...... 58 5.3.2 Wave Tank Experiments...... 58 5.3.3 Surface Water Profiling ...... 61 5.3.4 Subsurface Water Column and Dispersant Effectiveness...... 64 6.0 Conclusions...... 73 7 .0 Research Initiatives...... 7 6 7 .1 Identified Knowledge Gaps from the Literature Review ...... 76 7 .2 Gaps Addressed Under This Research Initiative ...... 7 6 7 .3 Gaps Identified from These Studies and Future Research Initiatives ...... 77 7.3.1 Laboratory Dilbit Fate and Behaviour Studies ...... 78 7.3.2 Meso-Scale Studies and In-situ Field Work ...... 78 8.0 References ...... 80

006 List of Figures

Figure 1-1 Map of the Athabasca, Cold Lake and Peace River oil sands in Alberta ...... 9 Figure 1-2 Illustration of the proposed Enbridge Northern Gateway Pipeline and tanker routes ....10 Figure 2-1 Simulated distillation data of oil sands products...... 13 Figure 2-2 The effects of regular waves, breaking waves and water temperature on dispersion of IFO180 ...... 23 Figure 3-1 Evaporative weathering of oil ...... 26 Figure 3-2 Density and viscosity at 0° C and l5°C as a function of evaporative weathering for AWB and CLB ...... 31 Figure 3-3 Pan evaporation results for two dilbit oils and IFO 180 ...... 33 Figure 3-4 High temperature simulated distillation plots ...... 35 Figure 3-5 Changes to AWB, CLB Hydrocarbon Group Composition with evaporative weathering, compared to conventional fuels ...... 37 Figure 3-6 n-Alkane distributions for AWB...... 39 Figure 3-7 Biomarker distributions for AWB...... 39 Figure 3-8 PAH distributions for AWB...... 41 Figure 4-1 Effects of evaporative weathering changes on density as a function of temperature for AWB and CLB samples ...... 46 Figure 4-2 Cold Lake Dilbit slick on saltwater, post-photo-oxidation...... 48 Figure 4-3 Densities of AWB samples before and after mixing with saltwater...... 49 Figure 4-4 Mixtures of CLB and kaolin sediment in saltwater ...... 51 Figure 4-5 Mixture of the W3 fraction of AWB and kaolin sediment ...... 52 Figure 4-6 Mixture of the Fresh and W2 fractions of AWB and diatomaceous earth sediment after 24 hours of settling ...... 53 Figure 4-7 Mixture of the Fresh and W2 fractions of CLB and sand after 24 hours of settling ...... 53 Figure 5-1 A collage of photos of wave tank studies ...... 60 Figure 5-2 Hydrocarbon concentrations (GC-FID) in surface samples collected during the treatment of CLB under breaking waves in the wave tank ...... 61 Figure 5-3 Saturates, aromatics, resins and asphaltenes in extracts of surface samples collected during wave tank studies of CLB ...... 62 Figure 5-4 IPAH concentrations in surface samples collected during various oil spill treatments of AWB ...... 63 Figure 5-5 IPAH concentrations in surface samples collect during various oil spill treatments of CLB...... 63 Figure 5-6 Cumulative particle size distributions for AWB and CLB fromLISST#l ...... 66 Figure 5-7 Cumulative particle size distributions for CLB from LISST#2 ...... 67 Figure 5-8 Hydrocarbon concentrations in breaking wave conditions at three depths and the effluent for CLB, CLB with dispersant, CLB with MF and CLB with dispersant and MF ...... 68 Figure 5-9 Interfacial tension for mesocosm testing of CLB under various treatment conditions .... 70 Figure 5-10 Dynamic dispersion effectiveness for AWB and CLB in breaking wave conditions using various treatment options ...... 71

007 List of Tables

Table 2-1 Selected Physical Properties and Chemical Data for Diluted Bitumen Products ...... 13 Table 2-2 Selected Physical Properties of Intermediate and Heavy Fuel Oil ...... 18 Table 3-1 Physical Properties of AWB ...... 29 Table 3-2 Physical Properties ofCLB ...... 30 Table 4-1 Linear regressions for density as a function of temperature for measured oils ...... 4 7 Table 4-2 Physical fate and buoyancy of oil-saltwater and oil-saltwater-sediment mixtures ...... 50 Table 5-1 Wave tank experimental conditions ...... 59 Table 5-2 Analysis of variance comparing the effectsof dispersant, mineral fines and mineral fines/dispersant to the natural dispersion of CLB under breaking waves ...... 72 Table 5-3 Analysis of variance comparing the effectsof dispersant, mineral fines and mineral fines/ dispersant to the natural dispersion of AWB under breaking waves...... 72

008 Executive Summary

Effective spill response depends on good scientific understanding of petroleum product behaviour in the environment ( e.g., movement and changes in physical properties and chemical composition of the oil). This study reports the early research simulating diluted bitumen products spilled at sea. This work was undertaken by the Government of Canada as part of the first phase of a strategy to implement a world class prevention, preparedness and response regime for oil spills from ships.

The behaviour of the diluted bitumen products was studied under laboratory conditions in three phases. First, the properties and composition of two samples representative of products currently being shipped in Canada were measured before (fresh) and after (weathered) exposure to environmental conditions. Secondly, the potential for evaporation, exposure to light, mixing with saltwater, and sediments in the saltwater to affect whether diluted bitumen products float or sink in saltwater was examined. Finally, the effectiveness of two existing spill treating agents meant to disperse spilled oil products was evaluated.

This work is a collaborative effort between the Emergencies Science and Technology Section, Environment Canada; the Centre for Offshore Oil, Gas and Energy Research, Fisheries and Oceans Canada; and CanmetENERGY, Natural Resources Canada. As well as the laboratory and wave-tank experiments, a literature review was conducted to identify knowledge gaps on the physical and chemical properties of conventional and non-conventional heavy oils, andtheir fate and behaviour in marine environments. The use and effectiveness of oil spill treating agents is also reviewed for heavy oils.

Two diluted bitumen products, Access Western Blend (AWB) and Cold Lake Blend (CLB), were selected for study as the highest-volume products transported by pipeline in Canada for 2012-2013. The physical characteristics and chemical composition of each product were measured to aid in potential spill preparation and response.

The major results of the studies were: • Like conventional crude oil, both diluted bitumen products floated on saltwater (free of sediment), even after evaporation and exposure to light and mixing with water; • When fine sediments were suspended in the saltwater, high-energy wave action mixed the sediments with the diluted bitumen, causing the mixture to sink or be dispersed as 1 floating tarballs ;

1 The use of the term "tarball" in this report followsconvention in the literature and refers to the consistency of floating, heavily-weathered oil. It does not describe the chemical composition of the product.

009 • Under conditions simulating breaking waves, where chemical dispersants have proven effective with conventional crude oils, a commercial chemical dispersant (Corexit 9500) had quite limited effectiveness in dispersing dilbit; • Application of fine sediments to floating diluted bitumen was not effective in helping to disperse the products; • The two diluted bitumen products display some of the same behaviours as conventional petroleum products (i.e. fuel oils and conventional crude oils), but also significant differences, notably for the rate and extent of evaporation.

Beyond informing subsequent studies, these results will also immediately help inform spill responders and computer models to better understand and predict the fate and behaviour of these non-conventional petroleum products in the marine environment. Better understanding the potential impacts of a non-conventional oil spill on the marine ecosystem, assists decision makers in making appropriate spill response and remediation choices. This suite of scientific research and activities will advance the knowledge of non-convention petroleum products; provide a better understanding of the potential consequences of a spill of diluted bitumen petroleum product on the marine ecosystem, and information to assist in response and remediation efforts.

010 1.0 Introduction

The Canadian oil sands contain the world's third-largest oil reserves after Saudi Arabia and Venezuela ( Alberta Energy, 2011 ). A map of the source oil sands deposits in northern Alberta can be found in Figure 1-1. Increased exports of oil sands products have been proposed by industry, involving pipeline, rail and marine tanker transport. These recent proposals include the Enbridge Northern Gateway pipeline to Kitimat, British Columbia (proposed pipeline and tanker routes shown in Figure 1-2), the Kinder Morgan Trans Mountain pipeline expansion to Burnaby, B.C., as well as pipeline projects contemplated for eastern Canada by TransCanada Corporation and Enbridge.

Bitumen is produced from the natural oil sands deposits by a number of processes, including direct mining and in-place extraction (Read and Whiteoak, 2003). The produced raw bitumen is a semi-liquid material at room temperature, and is too viscous to transport through a pipeline. In order to move it to market by pipeline, bitumen is diluted with either condensate or synthetic crude oil to form "dilbit" with viscosity, density and other properties engineered for pipeline transportation and use by the customer refineries. In many of the new projects being contemplated, diluted bitumen would be transported by pipeline to a coastal port, and then be shipped by marine tanker to foreign markets.

Thepotential range of behaviour, fate and treatment options for a possible marine spill of diluted bitumen products is not well understood. While some information on the physical properties and chemical compositions are available from scientific literature and industry sources, there is little information on the spill behaviour, fate, impacts and remediation options for diluted bitumen spills. The behaviour of dilbit in the marine environment, andits potential for alteration by evaporation, solar exposure, or mixing with water and sediment, are largely unknown. The conditions in which diluted bitumen products may float or sink in marine conditions have been the subject of much debate. In addition, the effectiveness of spill treating agents including dispersants and other treatments is not known.

In response to these knowledge gaps and concerns raised by these uncertainties, the Canadian federal government announced the World Class Tanker Safety System on March 18, 2013. Along with a number of tanker safety measures under this program, the Minister of Natural Resources Mr. Oliver, announced that the government will conduct scientific research on non-conventional petroleum products, such as diluted bitumen, to enhance the understanding of these substances and how they behave when spilled in the marine environment (Transport Canada, 2013).

This initiative mandates the Department of Fisheries and Oceans, Environment Canada, and Natural Resources Canada to conduct scientific research and activities on the behaviour, fate and appropriate response to possible marine spills of non-conventional petroleum products. The overall goal of this proposed research program is to characterize the chemical and physical 7

011 behaviour of spilled unconventional oil products in order to provide scientific support on assessing the risks and potential impacts a spill might have on the environment and to guide responders in deploying effective response and remediation approaches. The results of this research will also help in the development of tools such as predictive models on the fate and behaviour of spilled non-conventional petroleum products in the marine environment.

The first phase of the Government of Canada's scientific research initiative is planned to continue through 2016. However, in the interests of providing timely information in event of a spill and to guide future directions for research on marine spills of diluted bitumen, the three government departments have conducted a preliminary study on two selected diluted bitumen product blends: Access Western Blend and Cold Lake Blend. Thesetwo products, obtained from industry, were selected to represent the highest volume dilbit products transported by pipeline in Canada in 2012-13. The objectives of the present preliminary research activities are to:

• Conduct a literature review of information available in the public domain about the behaviour and treatment of diluted bitumen spills (Chapter 2); • Provide baseline data on physical characteristics and chemical composition of the two oil sands products (Chapter 3); • Conduct laboratory-scale studies on weathering and buoyancy of the products according to a matrix of environmental conditions which are high but of the same order of magnitude of measured levels in Canadian coastal environments. (Chapter 4); • Conduct meso-scale wave tank studies to determine the fateand transport of spilled dilbit products, and the effectiveness of the application of oil spill countermeasures, under various sea states in cold water conditions (Chapter 5).

Together, this suite of scientific research activities will advance the knowledge of diluted bitumen petroleum products composition, fate and behaviour in marine environments. The data and findingsof the present work will improve the preparedness and response for marine spills of non-conventional diluted bitumen products. Responders will be able to make informed decisions on the appropriate oil spill response options and strategies, including the possible use of spill treating agents.

012 Wood8Uffato NaUonalPark

Grande Prairie•

Edmonto!:ir' .,.,.,_,.... • UNallOMIPart!.11

BitumenP.y 'Calgary Thickness (metres) -60-100 -50-60 -•0-50 l::=130-40 Medicine c=l20-JO Hat,. c=J 10-20 100 200km c:::::I1.5-10 50 llr,lethbrldge

www.ags.gov.ab.ca

Figure 1-1. Map of the Athabasca, Cold Lake and Peace River oil sands in Alberta (Alberta Geological Survey, 2013).

013 N U.S.A A CANADA

...... ' ,. T PrtnceRupert I 'Jl(e_ • • Skeena _, I I 'I ' """"""""' I •

I I I

Queen Charto� Islands

un, Tlrlnet CJTerrestrial PDA con.�ed Channei As.ses.&men1Area - NorthernApproacn • • No.rtnemApproaCI\Lane - SOOtnem",)pro:teh soumemApproadl - (ViaPllncrpe Channel) • • sou:nemA;)proacn Lan.e t'bPnndpe Channet) � lntiemairon.¥6order

JWA-1048334-2746

Figure 1-2. Ilhlstration of the proposed Enbridge Northern Gateway Pipeline and tanker routes (NEB, 2012)

014 2.0 Literature Review of Diluted Bitumen and Synthetic Bitumen Spills, Behaviour and Use of Chemical Countermeasures

A comprehensive search for relevant literature available in the public domain was conducted. The primary focus of this search was to determine what informationexists on: a) the properties and composition of diluted bitumen as they relate to spill behaviour; b) information on the behaviour and fate of diluted bitumens and similar products; c) the use of chemical dispersants to treat diluted bitumen (dilbit) and synthetic crude oil-bitumen (synbit) mixtures and similar petroleum products; and d) to understand the state of knowledge about the influence of dispersants on the fate and transport ofdiluted bitumen products following a spill in the marine environment. Anticipating that limited information would be available on dilbit and synbit, the literature search also included non-conventional crude oils such as Orimulsion and fuel oils of a grade that might be similar to the products of particular interest.

Sources used to conduct the search were primarily from the Internet: Elsevier Scopus, Google, Google Scholar, Web of Science, JSTOR and Science Direct. The Department of Fisheries and Ocean's "Waves" search engine, the Environment Canada Arctic and Marine Oilspill Program Technical Seminar proceedings and the website of the International Oil Spill Conference were also used.

2.1 Properties and Composition of Diluted Bitumen Products from the Canadian Oil Sands

2.1.1 Diluted Bitumen Products

Oil sands bitumen is the oil found in a mixture of sands and clays native to the north eastern comer of Alberta. Over a geological timeframe, the lighter, more water soluble components of the oil were biodegraded when the oil came close to the surface of the earth and so cooled enough to allow microbial activity. The remaining oil has relatively high viscosity, organic acid and sulphur content. Natural bitumen is defined as a petroleum with a gas-free viscosity greater than 10 000 cSt compared to an extra-heavy crude with viscosity <10 000 cSt at reservoir temperatures (WEC, 2010). By this definition, oil sands bitumen can be a mixture of bitumen and extra heavy oil (ERCB, 2013). Unconventional techniques are required to produce the oil, including mining, if the ore is close to the smface, or heating with steam, if the reserve is deep. As produced, bitumen does not meet transmission pipeline specifications for density and viscosity.

For transportation purposes, lighter oil called diluent is added to bitumen to decrease its viscosity and density (Crosby et al., 2013). The most commonly used diluent is naphtha-based oil called condensate that can include by-products of natural gas. Typically, dilbit consists of blends of 20% to 30% condensate and 70% to 80% bitumen (Crosby et al., 2013; Crude Quality, 2013). As an alternative to condensate, mixtures of synthetic crude oil and bitumen are also 11

015 blended at approximately 1: 1 ratios and are known as "synbit." Synthetic crude oil is upgraded bitumen, produced by either coking or residue hydrocracking the bitumen, followed by hydrotreating to crack the larger molecules into smaller species and remove sulphur (Crosby et al., 2013). Bitumen producers create a variety of diluted bitumen products for transport.

Information on the chemical compositions and physical properties of oil sands bitumen and the blended bitumen products is essential in order to determine their fate and behaviour if spilled in the environment. The crudemonitor.ca data portal maintained by Crude Quality Inc. (Crude Quality, 2013) has detailed analyses of selected physical bulk properties and chemical constituents of the oils shipped in pipelines from Alberta, with data covering the last five years. Table 2-1 presents the annual averages and the standard deviation in those averages measured for diluted bitumen and other related oil sands products between October 2012 and October 2013. Figure 2-1 shows the average simulated distillations for the same rangeof products.

A range of variability can be seen in the annual density averages of approximately 0. 5% for all products. For other properties, however, including chemical compositions, larger ranges of variation canbe seen in the single year averages for sulphur and sediment content, as well as the total abundance of light ends (the sum of all butanes to decanes) and in the aromatic benzene, toluene, ethylbenzene and xylenes (BTEX) abundances. As expected, those products made from condensate generally can be seen to have higher levels of light ends and BTEX compounds.

The simulated distillation curves, from Figure 2-1, indicate a moderate range of variability in the diluted bitumen products constituted from condensate (Figure 2-l(A)). Much larger differences, however, are apparent in the simulated distillation curves for products diluted with synthetic crude oil and other petroleum products (Figure 2-l(B)).

While the crudemonitor.ca data portal provides an excellent source of data to the dilbit production and transportation industries and their customers, for the purposes of potential spill planning and response, several physical and chemical parameters are not included in the datasets. Parameters important for spill planning and response include: viscosity, surface and interfacial tensions, as well as alkylated families of polycyclic aromatic hydrocarbons. In addition, it is very desirable to measure not only the fresh product as produced or in the pipeline, but also product that has been evaporated and otherwise exposed to environmental weathering processes, to better understand the evolution of a spill.

016 Table 2-1. Selected physical properties and chemical data for diluted bitumen products (Crude Quality, 2013). Name Product Density Sulphur Sediment LightEnds* BTEX 3 (kg/m ) (wt%) (ppmw) (vol%) (vol%) Condensate Blends Access WesternBlend (AWB) 922.9± 4.6 3.94 ± 0.09 89 ± 8 24.1 ± 1.7 1.20± 0.15 Borealis Heavy Blend (BHB) 927.4 ± 5.2 3.67 ± 0.29 94± 27 24.1 ± 1.7 0.99± 0.09 ChristinaDilbit Blend (CDB) 924.9± 5.2 3.88 ± 0.09 88± 41 22.8 ± 2.2 1.12± 0.17 Cold Lake (CL) 927.7 ± 5.0 3.78 ± 0.08 94± 42 20.4 ± 1.5 1.06± 0.17 Peace River Heavy (PH) 930.5 ± 4.7 5± 0.1 97± 30 22.4 ± 1.1 1.02± 0.09 Statoil Cheecharn Blend (SCB) 928.8± 4.5 3.81± 0.09 169± 99 24.1 ± 2.3 1.06± 0.14 Western Canadian Select (WCS) 928.1 ± 4.3 3.50 ±0.07 284 ± 23 18.3 ± 1.3 0.83± 0.12 Blends Other than Condensate Borealis Heavy Blend (BHB) 927.4 ± 5.2 3.67 ± 0.29 94± 27 24 ± 1.7 0.99± 0.09 Statoil CheecharnBlend (SCB) 928.8 ± 4.5 3.81 ± 0.09 169± 99 24.1± 2.3 1.06± 0.13 Long Lake Heavy (PSH) 932.6± 3.6 3.21± 0.16 18 15.9± 1.2 0.94 ± 0.10 Statoil Cheecharn Synbit (SCS) 930.5 ± 4.2 3 07 ± 0.09 71± 11 13.4± 1.3 0.76± 0.09 Surmont Heavy Blend (SHB) 936.1 ± 3.8 3.08 ± 0.11 101± 42 11.3 ± 0.9 0.59± 0.09 Suncor Synthetic H (OSH) 936.5 ± 2.2 3.07 ± 0.09 39 10.4 ± 1.0 0.44 ± 0.08 Albian Heavy Synthetic (AHS) 938.7 ± 3.5 2.46 ± 0.23 784± 229 23.3± 1.4 0.94 ± 0.14 *Light End s compromise the sum of all butanes through decanes, inclusive.

800 800 (B) 700 700

600 600

'-.../ 500 500

Borealis Heavy Blmd Statoil Oieecham Synbit 200 Christina Dilbit Blend 200 Surmoot Heavy Blend Cold Lake Suncor Synthetic Heavy

Peace River Heavy Albian Heavy Synthetic 100 Statoil Cheecham Blend 100 Western Canadian Select 0 0 0 20 40 60 80 100 0 20 40 60 80 100 % Weight Off % Weight Off

Figure 2-1. Simulated distillation data of oil sands products. On Figure A (left), the distillation curves for seven types of bitumen diluted with condensate are plotted. On Figure B (right), the distillation curves for bitumen products diluted with other diluents are plotted. All data are 12-month averages fromOctober 2012 through 2013, from the crudemonitor.ca data portal (Crude Quality, 2013).

Data sets for MacKay Heavy Bitumen/Synthetic Crude Oil and Cold Lake Bitumen/ Condensate Blends were provided by Enbridge Northern Gateway to the NorthernGateway Joint Review Panel (Belore, 2010). Detailed chemical information, including alkyl polycyclic

017 aromatic hydrocarbon (P AH) data was later also provided in evidence (Stantec Consulting Ltd., 2012a). These data sets provide information on the fresh products as well as evaporated oils.

Detailed n-alkane and aromatic chemistry was reported by Yang and co-workers (Yang et al., 2011) for three oils sands bitumens and two oils sands diluted bitumens. They found that all three oil sands samples had similar compositions of n-alkanes, 2- to 5-ring alkylated P AHs, biomarker terpanes and steranes, and bicyclic sesquiterpanes. However, the composition of the diluted bitumens had been altered by blending with diluents or upgrading processing. A significant chromatographic feature of all of the oil sands products was the predominance of unresolved complex mixtures (UCM) in the chromatograms. The presence of a UCM indicates a complex mixture of hydrocarbons present in oil. This generally indicates both extensive biodegradation of the oil has occurred and that the existing hydrocarbon mixtures will be resistant to further biodegradation.

All of the bitumen samples showed evidence of prior biodegradation compared to a conventional crude oil. Normal alkanes and isoprenoid alkanes were almost completely absent in contrast to the make-up of a typical crude oil. For example, in total, the n-alkanes and isoprenoids make up 7.8% (w/w) of Alberta Sweet Mixed Blend crude oil (Environment Canada, 2006), a mid-weight crude oil. The bitumen samples also contained relatively low concentrations of 2-ring alkylated naphthalenes. The alkylated PAH compounds in all samples showed - increasing distribution profile of C0 < C1-< C2- < C3-< C4-, another marker of preferential biodegradation. In comparison, a typical crude oil has a "bell shaped" alkylated PAH distribution for each family of alkylated homologues. A typical distribution profile for a crude oil of fuel would be expected to have relative abundances for each alkyl PAH homologous family similar to

Co-< C1-< C2-;:::::C 3- > C4-.

All samples of bitumen and diluted bitumen reported by Yang and coworkers (Yang, et al., 2011) were shown to have abundant quantities of biomarker terpanes and bicyclic sesquiterpanes. These distinctive biomarker compositions indicate that they were generated from the same or very similar sources. Diamondoids were also found in significant abundance in samples of Alberta oil sands products. These biomarkers are all less susceptible to biodegradation than the P AH and n-alkane components and so may be useful oil forensic targets for environmental monitoring during spills, or during fate and behaviour experiments, including evaporation, dissolution and biodegradation studies.

018 2.2 Fate and Behaviour of Diluted Bitumen Products and Related Heavy Oils in the Marine Environment

2.2.1 Fate and Behaviour of Diluted Bitumen Products in Spills

A diluted bitumen blend spill occurred in 2007 from a pipeline operated by Kinder Morgan into Burrard Inlet, Burnaby, B.C. (TSB, 2007). Theproduct spilled was Albian Heavy, a blend of synthetic crude oil and heavier oil sands product. Approximately 224 cubic metres of oil were released, with 210 cubic metres being recovered (TSB, 2007). Oil escaped under pressure from a pipeline rupture. Spilled oil mi grated through the sewer system into Burrard Inlet where it began to spread on the water. Approximately 15 000 m of shoreline were affected by the spill.

An assessment of the spill clean-up and environmental impacts was commissioned by Kinder Morgan five years after the spill (Stantec Consulting Ltd., 2012b). The report of that assessment indicated that spill response operations were effective at removing oil from the environment and in limiting the short- and long-term effects of the spill. Oil was recovered by skimming and booming, as well as by flushing and removal from the affected shorelines.

Though shoreline intertidal zones were oiled, most marine sediments had only a small increase in measured PAH concentrations, with 20 of 78 monitored sites exceeding water quality guidelines (Stantec Consulting Ltd., 2012b). Levels of extractable hydrocarbons and PAHs for surface water quality requirements were met in 2007. Subtidal marine sediments were monitored through 2011, with most samples having levels of PAHs below the water quality requirements. Those subtidal sediment samples that did exceed the maximum regulated P AH levels appeared to be caused by sources other than the spill. Based on these observations, only trace amounts or less of oil from the 2007 spill appear to haveremained in the marine harboursediments.

A spill of dilbit occurred in 2010 from an Enbridge pipeline into the Talmadge Creek and Kalamazoo River, near Marshall, Michigan. An estimated 843 000 gallons (3190 m3) were released into the creek and river in flood conditions. Heavy rains caused the river to overtop existing dams and carried oil about 40 miles downstream on the Kalamazoo River. As of May 2013, the United States Environmental Protection Agency (EPA) estimated that 180 000 ± 100 000 gallons (680 ± 380 m3) of oil remained in the river as submerged oil (EPA, 2013).

An EPA report (Lee et al., 2012) shows epifluorescence micrographs of submerged oil particles in the Kalamazoo River. Lee and co-workers found that oil-sediment aggregates readily formed from native river sediments and the source diluted bitumen oils. These aggregates were stable over a 48-hour observation period. Oil-sediment aggregates similar to those made in the laboratory were also found in the river sediments, though at low concentration, presumed by the authors to be caused by the sediment agitation techniques being used forclean-up. These existing oil-sediment aggregates would have been approximately two years old at the time of the observations, indicating that dilbit can form stable aggregates with river sediments. 15

019 A third spill of Wabasca Heavy, a heavy oil/bitumen blended with condensate, occurred in 2013 near Mayflower, Arkansas, from an ExxonMobil pipeline into LakeConway. As of this writing, there is no reported study of the oil fatefor this spill.

SL Ross Environmental Research Ltd. (SL Ross, 2013) submitted a report on meso-scale weathering of Cold Lake Blend (CLB) diluted bitumen as evidence to the Enbridge Northern Gateway Joint Review Panel. In this report, the authors examined the behaviour of CLB, with a primary interest to "investigate the density change of the diluted bitumen product" as it weathered in the test tank for up to 15 days. The oil was recirculated in a "racetrack" style flume, while being exposed to shear mixing by passing through a water cascade once per cycle around the flume. A flow of air was also maintained above the water surface to simulate wind. Two test cycles were performed, one for 120 hours, the other for over 300 hours. The first test was conducted with limited light exposure; for the second, approximately one-quarter of each cycle in the flume was exposed to high intensity ultraviolet (UV) lamps. Both tests were conducted in fresh water at 15°C.

In both trials, the CLB oil rapidly formed a water-in-oil mixture. Without UV exposure, the oil/water mixture reached a maximum density of 0.995 g/mL with a water content initially rising to 39% (v/v) but falling to 25% by the end of the test. In the second trial, with UV exposure, the oil/water mixture density rose from the initial 0.945 g/mL to a maximum of 0.998 g/mL in 96 hours. Again, water content rose to a maximum of 34% within the first 24 hour period, but fell back to under 20% as the test proceeded. Viscosity of the oils in both tests increased continuously with time, reaching maxima of 200 000 mPa•s after 120 hours and over 1 000 000 mPa•s after311 hours for the firstand second tests, respectively.

The authors (SL Ross, 2013) remark that, at the water cascade, oil was observed to be shearing into small droplets (1-4 mm diameter). The oil mixed down into the water for a short time but quickly resurfaced to re-form a surface slick. As both tests proceeded and the oil viscosities increased, the oil droplet size at the weir was increased. In the test with UV exposure, after 48 hours small oil droplets were seen in the water column, with the water becoming cloudier with time. The authors concluded that this indicated a natural dispersion of oil droplets, which they estimated to be 30 to 70 µm in diameter by analysis of photographs. The authors also remark that this dispersion was not seen in the first, non-UV exposed test.

At the end of each test, oil was collected from the tank surface and the tank walls. Most of the oil was recovered from the tank surface, the remainder being 15% of the oil recovered from the tank walls deeper than 10 cm below the water surface. The authors attribute this to adhesion of neutrally buoyant drops in the water column to the walls of the tank. Though not stated in their discussion, this implies that at least 15% of the CLB dispersed into the water column over the course of their trials. The authors also note that no oil was found on the bottom of the tank.

020 2.2.2 Fate and BehaviourofOrimulsion in Spills

Orimulsion is a product manufactured until recently in Venezuela from Orinoco bitumen. Orimulsion consists of approximately 70% bitumen, 30% fresh water and a small amount of surfactant (<0.2%) to keep the bitumen in suspension during transport (Miller et al., 2001). Like dilbits and synbits, Orimulsion originates as extra heavy oil (bitumen). However, the Orimulsion is a bitumen-water emulsion that is stabilized with additives; the bitumen components can settle out of the mixture when left undisturbed (Jokuty et al., 1999). Analysis of two Orimulsion products (Wang and Fingas, 1996) can be found in the Environment Canada Oil Properties database (Environment Canada, 2006).

As Orimulsion was made from a similar bitumen resource to that used to blend dilbit, a review of Orimulsion fate and behaviour studies was conducted to draw any information from the similarity in composition of the source bitumen, while keeping in mind the large differences in diluents used for both.

In seawater, Orimulsion forms a cloud of dispersed bitumen droplets, and a dissolved PAH and surfactants aqueous phase, which quickly mixes into the receiving water body. The bitumen droplets chemically resemble a lightly weathered heavy fuel (Jokuty et al., 1999). Studies have shown that Orimulsion will float in salt water and sink in freshwater; however, behaviour in brackish water is more complex and depends on salinity, time, energy and temperature (Fingas et al., 2003; Fingas et al., 2005).

Orimulsion interactions with suspended sediments in water appear to be enhanced by high salinity, sediment surface area and high organic carbon content (Brown et al., 1995). Interestingly, Brown et al. (1995) report that the adhesion of heavy fuel oil (HFO) to the same sediments was negligible compared with Orimulsion, with the HFO taking up much less of the sediment (by mass) compared with the dispersed bitumen. In mid-range to high seawater salinity, dispersed Orimulsion bitumen formed aggregateswith sediments that sank easily. In high mixing energy studies, in seawater over 10% of the bitumen sank with 45 mg/L loads of suspended kaolinite (Stout, 1999). However, in fresh water, agglomeration and sinking were not observed, presumed by both authors to be caused by the action of the surfactant.

When stranded on shorelines, experiments showed (Harper and Kory, 1997) that, when fresh, Orimulsion behaves like a heavy oil (see Section 2.2.3 below) and readily penetrates into shoreline sediments. However, the authors reported that when weathered and/or dried, Orimulsion is "stickier" than fresh Orimulsion and will not penetrate beach sediments as easily. As it dries, Orimulsion forms a weathered bitumen coating on the surface of grains or pebble that is difficult to remobilize. These coatings can form quickly to depths of 4 cm in sand, 6-7 cm in granules and 7-8 cm in pebbles.

021 When fresh, dispersed bitumen particles were found to be more mobile within coarse grained beach substrate than a typical heavy fuel oil. Dispersed bitumen could be flushed from sediments, but normal tidal flushing was foundto be unlikely to remove it completely.

Additional studies were undertaken to evaluate bitumen adhesion and removal from the surface of rocks (Harper and Ward, 2003 ). It was found that wetting the surface of rocks was the important control mechanism of bitumen adhesion. Wetting of the rocks allowed for relatively easy removal of coalesced bitumen from the rock surface. Naturally occurring biofilms foundin the lower and mid-intertidal zones were also shown to reduce bitumen retention.

2.2.3 Fate and Behaviour of Heavy Oils in Spills

The possible fate and behaviour of diluted bitumen when spilled have been compared to those of lighter fuel oils in the early part of a spill, changing to heavier fuel oils as weathering occurs (ENG, 2011). The literature on intermediate and heavy fuel oils similar in density and viscosity to diluted bitumen products was therefore included in the present survey of the literature. The heavy oils considered at greatest risk for sinking are typically heavy fuel oils, sometimes referred to as Bunker C and Fuel Oil No. 6. Detailed analyses of several of these products can be found in the Environment Canada Oil Properties database (Environment Canada, 2006). Selected physical properties and chemical characteristics of representative intermediate and heavy fuel oils are presented in Table 2-2, taken from the Environment Canada database.

Table 2-2. Selected physical properties of Intermediate and Heavy Fuel Oil IFO 180 HFO6303 Pour Point (0C) 15 -1 Density (g/mL) 0°C 0.9794 1.0015 15°c 0.9664 0.9888 API Gravity 12.9 11.5 Dynamic Viscosity (mPa•s) 0°C 12400 241000 15°c 1920 22800 Emulsion Formation Stability Class Entrained Entrained Tendency and Stability Complex Modulus (Pa) 144 752 Water Content(% w/w) 42 57.7

Chemical Dispersability (SFT using Corexit 9500) <10% <10% Saturates (¾w/w) 49 42.5 Aromatics (¾w/w) 32 29.0 Resins (¾w/w) 15 15.5 Asphaltenes(¾w/w) 4 130 Waxes(¾w/w) 20 2.5

022 Michel (2010) recently reviewed the literature for heavy oils at risk of sinking in aquatic settings. The author also reviewed historical case studies when oil has been observed to sink. Michel notes that oils typically follow two paths to sinking, for oils less dense than water and for oils denser than water.

If the oil is less dense than water, then the oil will initially float, and will likely only begin to sink if it is mixed with a heavier material such as sediment. The author notes that sediment mixing can typically occur by two mechanisms: 1) stranding on a shore and mixing with sandy sediments; and 2) by mixing with sediments in the water column by wave action, away from shore. Michel notes that the oil can either take up just enough sediment to become neutrally buoyant in the water, and thus be transported by water currents, or be dense enough to sink directly to the water bottom. Finally, the author notes that oil has been observed to shed sediment over time, resulting in refloatation of the oil.

If the oil is more dense than water, then the oil will not float. In freshwater, these oils have a density greater than 1.0 g/mL (or API gravity less than 10); in seawater at 35 parts-per thousand the oil density must exceed 1.035 g/mL (an API gravity of less than 6). The author (Michel, 2010) notes that evaporation alone has been seen to be sufficient to cause sinking of oils that are already close to the density of water, to switch from the first sinking pathway to the second pathway.

The author also conducted a survey of spills of heavy oil which were reported to involve sunken oil. Several common characteristics were found in the 26 incidents reviewed:

• Of the spills where oil was denser then the receiving water and sank upon release, in all cases these were heavy refinedproducts or coal tar. Five were of "slurry oil" or heavy cycle gas oil. • The only crude oils reported to have sunk were heavy crudes from Venezuela. All sank only after the crude oil mixed with sediment from landing on beaches. • Fifteen of the 26 incidents involved heavy fuel oils, sometimes referred to as Bunker C or Fuel No. 6. Many of these were initially buoyant but sank after sediment mixing. • A few incidents involved unique or rarely shipped products including asphalt stock, carbon black feed stock and coal tar oil.

The author remarked that there were no examples in the literature of an inte1mediate fuel oil that had been observed to sink.

In a monograph compiled by the United States National Research Council (NRC, 1999), five majormechanisms for heavy oil sinking were proposed: evaporation; temperature variations with the day/night cycle causing cyclic sinking and refloating; uptake of solid matter, sand and similar granular material; photo-oxidation; and water uptake.

023 Long-term fates of fuel oil spills have been examined extensively in the cases of the Arrow (Owens, 2010) and the Prestige (Diez et al., 2007) spills. Additionally, the persistence of oils on shorelines, including several cases involving heavy fuels, was reviewed by Owens (2008).

The wreck of the Arrow spilled an estimated 55000 barrels (8700 m3) of Bunker C fuel, equivalent to the HFO described earlier, along 300 km of shoreline. The effects and clean-up of the spill were extensively studied with follow-up for30 years. The majority of the studies were unpublished, but are summarized and reviewed by Owens (2010).

Long-term persistence of the Arrow oil occurred by penetration and burial into rocky and cobble beaches primarily above the tidal zone. Following an initial period of removal and attenuation, oil chemistry was not found to change after being sheltered in the coarse sediment and cobble.

Following clean-up, much residue was left on the beaches as a band of oiled cobble. However, in the three years following, it was found that 75% of the oiled cobble shorelines had eroded even in areas of low wave and tidal action. The formation of an oil-water-clay emulsion made the oil bioavailable and removable by the surf. Twenty years following the spill, less than 13 km of oiled shoreline remained, much of the affected shores having been cleaned by this natural process. Oiled cobble was continuing to be slowly eroded by surf, being removed as it was slowly exposed.

Owens (2008) identifies six conditions for long-term persistence of oil on shorelines: (i) heavy or highly-weathered oil types; (ii) large volumes of stranded oil; (iii) coarse-grained surface sediments with a penetration-limiting layer below; (iv) deposition above normal wave action; (v) penetration or burial to a depth below the layer of mobile sediments; and (vi) non eroding shorelines. Under these conditions, oil can persist buried in the sediment, sequestered in coarse sediments, or form a weathered and degraded asphalt pavement above the tidal zone.

In the case of the Prestige spill in 2002, the fate of the heavy fuel oil at sea and on land was examined (Diez et al., 2007). Evaporation of the oil was found to be a slow, "steady depletion" over the course of a year. Oil on beaches and rocky shores degraded more quickly than in open seawater, thought by the authors to be caused by a combination of higher temperatures on shore and enhanced biodegradation by beach organisms. Dissolution was found to account for a net loss of less than 2% of the total oil, but removed up to 30% of the more soluble P AH species. Over the first year of the spill history, the authors found that biodegradation accounted for a minor part of the oil weathering. Similarly, changes to the oil chemistry by photo-oxidation were "not significant at sea" during the year following the spill. The authors conclude that for the first year following the spill, the heavy fuel oil was highly persistent, with "very low incidence of the natural weathering process."

024 Recently, highly degraded tarballs on the west coast of Spain have been linked to the wreck of the Prestige (Bernabeu et al., 2013), a decade after the event. The authors conclude, based on the chemical signatures of degradation, that the tarballs likely originate from removal of older oil deposited in subtidal sediments. That is, the tarballs originated in sunken oil that has become freeof ocean sediment and refloated as largeballs and patties.

2.3 Use of Dispersants on Potential Spills of Diluted Bitumen and Related HeavyOils

2.3.1 Use of Dispersantson a Diluted Bitumen Spill

Very fewstudies have been conducted on the use of chemical dispersants on dilbit, syn bit or condensates, and there is no record of chemical oil dispersant being used on an actual dilbit or synbit spill in the ocean. The efficacy of chemical dispersants when used on a surface diluted bitumen slick is largely unknown. In addition, the chemical composition of dilbit can differ depending on the bitumen that is used, and the solvent that is used to dilute it (Crude Quality, 2013); therefore, selection of an ideal dispersant to treat a diluted bitumen spill may prove challenging. In the absence of data specific to dilbit and synbit, the application of dispersants to spills of some conventional hydrocarbon products such as heavy fuel oils (which are characterized by a high proportion of relatively heavy asphaltenes) may in part be comparable.

2.3.2 Use of Dispersantson Orimulsion

The use of chemical dispersants on spills of Orimulsion is not well studied; however, research has shown that Orimulsion will disperse naturally in freshwater. In seawater, following initial dispersion, a certain amount of coalescence of the bitumen droplets will occur, eventually forming a surface slick (Ostazeski et al., 1997). Flume studies with seawater demonstrated that the coalescence of the bitumen occurs rapidly, and bitumen droplets were observed to increase in size from a mean droplet diameter of 120 µm to 140 µm after 8 hours (Ostazeski et al., 1997). Coalescence of oil following dispersant application would suggest that the dispersant product was ineffective in the dispersion of Orimulsion. The applicability of these results to the use of dispersants on spills of dilbit is unclear, based on the vastly different formulations of the two products.

2.3.3 Use of Dispersantson Heavy Fuel Oils

A recent review by Chapman and co-authors (Chapman et al., 2007) examined the use of dispersants on high viscosity oils similar in density and viscosity to diluted bitumen. The authors note that, "as a general rule of thumb, oils with high viscosities and pour points tend to be much less dispersible as delivery of the surfactant to the oil-water interface becomes more difficult." In reviewing a case study of the Natuna Sea spill of crude oil, they note that when the crude oil viscosity exceeded 50 000 mPa, by reduced temperature and environmental weathering,

025 dispersant use ceased to be effective. Chapman et al. (2007) do note effective uses ofdispersant on higher viscosity fuels, including IFO 380, but only in the warm water conditions of the Helicom spill. Chapman et al. (2007), also remark that high asphaltene and wax contents tend to further reduce dispersant effectiveness.

The Centre for Offshore Oil, Gas and Energy Research has conducted wave tank oil spill dispersion experiments on unweathered intermediate fuel oil, IFO 180 (Li et al., 2010). Figure 2- 2 shows that IFO 180 is considerably more difficult to disperse. Looking more closely at Figure 2-2, IFO 180 seems to have been dispersible in breaking waves when the temperature was >15°C, but showed very low dispersibility in non-breaking waves or temperatures below 10°C (Li et al., 2010). In the study ofLi et al. (2010), analysis of covariance indicates the wave type and temperature si gnificantly affected (p>0.05) the dynamic dispersant effectiveness. Similar results for the dispersibility of IFO 180 have been found in field trials conducted by Colcomb et al. (2005), where chemical dispersants were shown to be effective at a surface temperature of 15° C and with winds of7-14 knots. Small-scale laboratory dispersant experiments (Clark et al., 2005) further corroborate these reports by showing that IFO 180 is dispersible with chemical oil dispersant, provided that a high level of mixing energy is present. The literature on IFO 180 suggests promising dispersibility characteristics for these two conventional oils; however, the comparability of the IFO products and typical dilbits is complicated by their differing chemistries. As described earlier, dilbit has been compared to much heavier conventional oils such as IFO 380, Bunker C or No. 6 Fuel Oil, which are all considerably heavier than IFO 180. These fuels have API gravities in the range of11. 4-14.1 (Environment Canada, 2006). Lune! and Davies (2001) and Chapman et al. (2007) suggest that chemical oil dispersants are not a particularly effective treatment option for spills ofheavy bunker oil or IFO 380. However, Lune! and Davies (2001) also explain that under favourable conditions, some varieties of IFO 380 are dispersible with dispersant products such as Corexit 9500, Dasie Slickgone LTSW, Inipol 90 or Superdispersant 25. The in-situ viscosity of the oil is the primary factor that dictates how dispersible an IFO 380 product will be. All other considerations aside, in ocean temperatures below 10°C or in cases where the oil has formed a highly viscous oil-water emulsion, chemical oil dispersants are not likely to be effective (Lune! and Davies, 2001). In addition, sea state can have both a positive and a negative impact on the dispersability ofIFO 380.

026 (A) 4 (B) 4 Water Water 9 R, 11.6 -s, .9 - 3 •• ..,.. B, ····•··· R, 13.5 .s 12.1 C _.,_ B, 13.8 ---- R, 16.1 -+- B, 2 17.6 C ... Q) ...... ··"' (J !---- .. .,,.. .. 1 J.■ _ � ::t::w.,_ .. _ • .,.,; .. .. 8 �•··...... ,.___. -----··- ...... , 4 • 0 ...,,...... "T't �, --- ��-+-'-�'-+-'�-'-+-�--'-+�...... f-'-->-�+-' +-'-�-+-'-�'-+-'�-'--+-�--'-+�...... ��+-' 0 0 0 0 10 20 30 40 50 60 10 20 30 40 50 60 Time (min) Time (min) C �------( ) 4 (D) 40 ♦ \ Corexit 9500 ,, r Corexit 9500

I 9 R, 10.1 -- B, .8 - •··•·· B, 10.2 ....,. ... R, 10.5 ;. \ �C 20 :/ \ -a-- B, 14.2 ---- R, 15.5 \ ' (I -+• B, 15.6 i • ' 8 10 t 6 I·.------...... 0 ', ...... __ _ J' ·•.. �------, ' -.. •• � ��.-- 0 0 .J-4,GIW""""�J. _ , """'"....,."-�=:..i:;:=�=:.::ct\t-J.- ---� 0 10 20 30 40 50 60 0 10 20 30 40 50 60 lime(min) Time (min)

(E) 4 (F) 40 �------� SPC 1000 SPC 1000 R, 12.1 ...... R, 12.7 9 9 ---- R, 16.8 -s, . _g: 2 �C 20 •••,,. •• B, 13.1 -•- B, 15.6 C 9 -- __,.,_ B, 16. §0 1 _.__ ------,,,.. - 8 10 6 ... · ..-:-;...... -.,.: ••···• ..,..,""�-··••··· ··•····· -- 0 • • 0 � 0 10 20 30 40 50 60 10 20 30 40 50 60 Time (min) Time (min)

Figure 2-2. The effects of regular waves (panels A, C and E), breaking waves (panelsB, D and F) and water temperature (numbers in the legend) on dispersed IFO180oil concentration(1 mg/L = 1 ppm) in the middle of the water column and10 m downstream from the initial oil slick using two chemical oil dispersants. R = regular, B =breaking.Figure taken from Li et al., 2010.

Although an abundance of wave energy will help with the physical mixing and dispersion of the oil, it is also likely to accelerate the production of oil-water emulsion, ultimately decreasing the effectiveness of chemical oil dispe1-sants (L unel and Davies, 2001 ). In situations of high wave energy, the window of opportunityfor use of chemical oil dispersants is likely to be narrow, i.e., the timeframe during which the counte1measure is effective (Lunel and Davies, 2001).

In small-scale laboratory testing, Stevens and Roberts (2003) and Clark et al. (2005) have also found that given the right conditions, chemical oil dispersants can be effective on IFO 380, 23

027 albeit at a lower dispersion efficiency than would be achieved with lighter crude oils. Using the Warren Spring Laboratory Test, Stevens and Roberts (2003) found that seven of the nine IFO 380 oils they tested could be dispersed at 15°C using a dispersant-to-oil ratio (DOR) of 1:25. Corexit 9500 and Slickgone EW were the most effective dispersants used in their tests. Similarly, Clark et al. (2005) showed that IFO 380 did disperse in the BaffiedFlask Test, Exxon Dispersant Effectiveness Test and the Warren Spring Laboratory Test, but did not disperse as well using the low energy Swirling Flask Test. In these tests, Corexit 9500 and Superdispersant 25 proved to be most effective. Expanding further on their results, Clark et al. (2005) suggest that mixing energy was the most predominant factor in determining dispersion efficiency, and that decreasing the DOR from 1:50 to 1:25 made a bigger difference to IFO 380 than to a relatively lighter oil, IFO 180.

Larger-scale field trials conducted by Colcomb et al. (2005) to examine the dispersion of IFO 380 also showed mixed results. In their study, IFO 380 was both slow to disperse, and ultimately only partially dispersible in sea temperatures of 15°C and with wind speeds between 7 and 14 knots (Colcomb et al., 2005). These experiments support the findings of Clark et al. (2005) in that more wind (higher wave energy) resulted in better dispersion, and that a DOR of 1:25 proved to be more effective than a DOR of 1:50 when attempting to disperse IFO 380. The results of these field tests were later replicated and confirmed at the OHMSETT facility in New Jersey (Trudel, Belore, Guarino, Lewis, and Mullin, 2005), but unfortunately, neither Colcomb et al. (2005) nor Trudel et al. (2005) investigated the potentially competing influence of wave energy on the dispersion ofIFO 380 that was identified by Lune! and Davies (2001).

028 3.0 Physical Characteristics and Chemical Composition of Diluted Bitumen Products

3.1 Samples

In consultation with industry stakeholders, two diluted bitumen products were selected for study on the basis of production and transportation volume: Access Western Blend (AWB) and Cold Lake Blend (CLB). As these products were collected in mid-winter, they contained the maximum content of diluent needed to meet viscosity requirements for transmission pipeline transportation (Crude Quality, 2013).

For comparison, an intermediate fuel oil, IFO 180, used for marine ship fuel, and a mid to-light reference crude oil, Alberta Sweet Mix Blend, reference pour #5, were also included in some tests. 3.2 Methods

3.2.1 Evaporative Weathering

For most of the measurements reported in this section and for the buoyancy tests reported in Chapter 4, environmental weathering was simulated by rotary evaporation. The oil weathering system consists of a Buchi R220 rotary evaporator with a flask rotation speed of 135 revolutions per minute (rpm) and a positive air flowof 13 L/min (Figure 3-1). The samples are held at 80°C. At set intervals, the sample flask is removed and weighed to monitor mass loss. The system has been demonstrated (Fieldhouse et al., 2010) to result in chemical compositions identical to those obtained from evaporation from an open pan for the same total loss of oil mass. That is, composition of oil evaporated in the rotary evaporator is identical chemically to oil evaporating in an uncontrolled pool at ambient temperature and pressure.

Typically, three weathered fractions were prepared for each oil sample. The initial weathering period was 48 hours, the duration chosen to simulate a heavily-weathered state of an oil in the environment. In addition, intermediate fractions of approximately one- and two-thirds of the 48-hour loss by weight were prepared. These were designated as Wl, W2 and W3 to represent increasing degrees of weathering (Hollebone, 2010). For the present work, an additional highly weathered fraction, W 4, was also prepared at 96 hours' evaporation at 80°C.

For the wave tank studies reported in Chapter 5, the AWB and CLB were artificially weathered 8.8 and 6.2%, respectively. This technique of weathering oil is consistent with the treatment of conventional oil prior to oil dispersion studies in the wave tank (Li et al., 2008).

029 Figure 3-1. Evaporative weathering of oil. (Photo: B. Fieldhouse, Environment Canada)

3.3 Characterization of Diluted Bitumen Products The physical properties of fresh and weathered dilbits were measured including viscosity (ThermoHaake VT550: Hollebone, 2010, and Brookfield LVT), density by AS1M D5002 (Anton Pm,· DMA 5000, Mettler Toledo DM45) and surface tension (using aKSV CAM 200 by the pendant/buoyant drop method: Cheng et al., 1990 and using aKriissKl1, by Wilhelmy plate method: Drelich et al., 2002). API gravity was determined from the density measurements by interpolation (Hollebone, 2010). Sulphw· content was measured by X-ray fluorescencefollowing AS1MD4294 . Pour point was measured according toAS1M D97.

The rate of oil evaporation from an open pan was determined following Fingas (Fingas, 2004). The boiling point fractions of the dilb its were determined by high temperature simulated distillation (AS1M D7169). The hyd rocaibon families of saturates, aromatics, resins and asphaltenes ("SARA" groups) were quantified by a combined GC-FID and solvent precipitation method (Hollebone et al., 2003), ooich is based on AS1M D2007, Appendix XL

Oil-water mixtw·es were prepared using a rotary end-over-end mixing of a 1:20 ratio of ° oil-to-water at a rate of several revolutions per minute for 12 hours at 15 C. This procedure has been shown to provide a high-energy mixing environment, with sufficient mixing time to ensure 26

030 that if a mixed water-oil state is possible, it will likely be reached during the test (Fingas and Fieldhouse, 2006). If a continuous oil-water phase is formed, the visco-elasticity and water contents are determined immediately after the mixing, followed by a second measurement seven days later.

The distributions of n-alkanes, PAHs and alky-P AH homologs and petroleum biomarker hopanes and steranes were characterized by gas chromatography coupled with mass spectrometry (Wang et al., 1994).

Unless otherwise noted in the text, all determinations for physical properties and chemical composition measurements were conducted in triplicate. Uncertainties are not shown on the tables below, but the error bars that appear on figures in Chapters 3 and 4 were determined from the standard deviation of the triplicate measurements.

Where applicable, measurement temperatures used were 0°C and l 5°C. This range covers most of the marine temperatures experienced in Canadian waters. On the Atlantic coast, monthly average surface water temperatures range from 0°C to 5°C north of Newfoundland to 0°C to 15°C in the Bay of Fundy (DFO, 2013a). In the Gulf of St. Lawrence, seasonal surface temperatures range from 0°C to l 7°C (Galbraith, et al., 2012). On the Pacific coast, a similar range of surface sea temperatures are experienced from 5°C to l 5°C in the Salish Sea to 5°C to 10°C in Hecate Strait (DFO, 2013b). For the behaviour and fate testing, including pan evaporation, emulsion formation in the present chapter and the sediment interactions reported in Chapter 4, 15 °C was chosen as the working temperature. This not only is a representative temperature for warm Canadian ocean waters, but also provides comparability with previous work on evaporation (Fingas, 2004), emulsion formation (Fingas and Fieldhouse, 2006) and behaviour studies (SL Ross, 2013).

3.4 Physical Properties of Diluted Bitumens

The measured physical characteristics and bulk properties for A WB and CLB diluted bitumen can be found in Table 3-1 and Table 3-2, respectively. The properties listed in these tables are amongthose most relevant to understanding and predicting the behaviour of oils when spilled. Density, viscosity and surface/interfacial tensions are considered among the most significant factors in understanding the behaviour of oil. Density governs the buoyancy of oil in water; viscosity, its rate of spreading and resistance to being dispersed into droplets; surface and interfacial tensions are important at smaller scales, governing the size of dispersed oil droplets, and the final thickness of a spreading oil film. API (American Petroleum Institute) gravity is a petroleum industry standard for describing the density of oil at 60°F (15.6°C) relative to that of water. Pour point is a measure of viscosity that defines the minimum temperature at which oil will flow. Flash point is the minimum temperature at which enough volatile chemicals evaporate to support ignition. Water and sulphur contents describe the percent concentration, by weight, of

031 water and elemental sulphur in the oil. Water is typically dispersed in the oil in droplets; sulphur is bound in a range of chemical forms from inorganic sulphides to organic sulphur compounds.

The tendency for emulsification, the formation of a water-in-oil colloid, is also a very useful test to interpret oil or petroleum behaviour in the environment. As described in section 3.3, following a 12-hour period of shaking, an oil-water mixture is assessed for emulsion formation. Measurements are made of the rheology of the oil, including the complex modulus, the resistance of the oil-water mixture to flow under dynamic (pulsed) stress. This quantity is a combination of both normal viscosity and elastic (reversible) deformation. Water content, as described above, is also measured. Measurements are taken both within 24 hours following mixing andagain seven days later to assess the stability of any colloids formed.

Based on these measurements, an oil-water mixture is classified into one of four states (Fingas and Fieldhouse, 2006): a stable emulsion, which can persist indefinitely; a meso-stable emulsion, which typically has a short half-life of hours to days; and an entrained-water state, which is typically lower-water content and complex modulus than the chemically-stabilized emulsions, but which still retains water forperiods of days to weeks. The last category is the no emulsion state, where oil and water remain apart. In the no-emulsion state, the viscosity of the oil is typically unchanged and the water content of the oil phase is minimal, following mixing.

032 Table 3-1. Physical properties of AWB. Degree of Evaporation (lV[ass Loss w/wo/o) Fresh Wl W2 W3 W4 (0%) (8.5%) (16.9%) (25.3%) (26.5%)

Sulphur Content(% w/w) 3.0 4.1 4.5 4.9 4.8 Water Content(% w/w) 1.5 0.9 0.2 00 00 Flash Point (0C) <-5 <-5 29 159 173 Pour Point( 0C) <-25 <-25 -6 24 33 Density(g/mL) 0° C 0.9399 0.9646 0.9949 1.02 14 1.02 11 15 °c 0.9253 0.9531 0.9846 1.01 27 1.0140 20°c 0.9148 0. 9547 API Gravity 20.9 16 .6 12 .0 8.2 8.0 Dynamic Viscosity 0° C 1.30E+03 9.82E+03 2.04E +05 9.35E+07 >l .00E +08 (mPa•s) 15 °c 347 l.72E+03 2.97E+04 2.52E+05 7.91E+06 40 °c 59.8 348 Emulsion Formation Stability Class Entrained Entrained Entrained Entrained DNF Tendency and Stability Complex Modulus 44 .6 89 .7 46 7 1.26E+04 NIA (Pa) Water Content 40 35 33 6 NIA (¾w/w) Surface Tension 0° C 31.2 31.9 NM NM NM (Air/Oil, mN/m) 15 °C 30.2 31 .1 31 .2 NM NM 20°c 27.5 Interfacial Tension 0° C 24.8 NM NM NM NM (Oil/Water, mN/m) 15 °C 24.2 28 .0 NM NM NM Interfacial Tension 0° C 25 .0 NM NM NM NM (Oil/33%0 Brine, mN/m) 15 °c 23 .8 26 .0 NM NM NM

NM - Not Measurable, too viscous; DNF - Did not Form, too viscous; NIA- Not applicable; *Measured at8.8 % evaporated via sparging.

033 Table 3-2. Physical properties of CLB. Degree of Evaporation(lV[ass Loss w/wo/o) Fresh Wl W2 W3 W4 (0%) (8.5%) (16.9%) (25.3%) (26.5%)

Sulphur Content(% w/w) 3.7 4.0 4.2 4.5 4.6 Water Content(% w/w) 0.6 0.2 0.0 0.0 00 Flash Point( 0C) <-5 -1 66 155 169 Pour Point( 0C) <-25 <-25 -12 21 27 Density(g/mL) 0° C 0.9376 0.9665 0.9909 10130 10176 15°c 0.9249 0.9537 0.9816 10034 10085 20°c 0.9216 0.9471 API Gravity 210 16.5 12.5 9.5 8.8 Dynamic Viscosity 0° C 803 6.98E+03 129E+05 185E+07 5.71E+07 (mPa•s) 15°c 285 1.33E+03 l.83E+04 3.91E+05 3.21E+05 40°c 59 175 Emulsion Formation Stability Class Meso Entrained Entrained Entrained Entrained Tendency and Stability Complex Modulus 90.5 64.0 267 3.60E+03 9.24E+03 (Pa) Water Content 72 46 38 9 10 (¾w/w) Surface Tension 0° C 30.0 30.8 NJ\,1 NJ\,1 NJ\,1 (Air/Oil, mN/m) 15°C 28.8 30.1 30.2 NJ\,1 NJ\,1 20°c 28.0 Interfacial Tension 0° C 30.6 NJ\,1 NJ\,1 NJ\,1 NJ\,1 (Oil/Water, mN/m) 15°C 27.7 28.9 NJ\,1 NJ\,1 NJ\,1 Interfacial Tension 0° C 30.4 NJ\,1 NJ\,1 NJ\,1 (Oil/33%0 Brine, mN/m) 15°c 26.3 27.6 NJ\,1 NJ\,1 NJ\,1

NJ\1- Not Measurable, too viscous

*Measured at 8.8% evaporated via sparging.

034 3.4.1 Properties of Diluted Bitumens

Many properties of the diluted bitumen products were found to change with increasing evaporation. Density, pour point, flash point and viscosity all increased with increasing evaporation. The changes in density and viscosity are plotted on Figure 3-2, with AWB on the left panel and CWB on the right. Densities are shown as circles, solid black at 0°C and open ° white circles at 15 C. Viscosities on both figures are similarly indicated by square points. In all cases, error bars based on the measured triplicate standard deviations are smaller than the size of the plotted points, and are thus not shown on Figure 3-3. In addition, the 0°C viscosity for the most weathered W 4 A WB samples was too high to be measured, and is not plotted on the figure.

1.04 IO" 1.04 IO" Access WesternBlend ColdLake Blend 1010 IO" 1.02 Density 0°C •• 1.02 Density 0°C • 0 • • ° 0 ° • 0 Density 15 C IO' �00 0 Density 15 C 0 IO' 00 ° ° 0 1.00 ■ Viscosity 0 C "' 1.00 ■ Viscosity 0 C p.,"' ° • ° □ Viscosity 15 C ■ 10' g D Viscosity 15 C • ■ IO' g 0.98 c 0.98 0 ■ c □ 107 ·ui IO' ·ui i 0 i 0 u u 00 00 ,q • lo' ,q • IO' � 0.96 > � 0.96 > "' 0 ■ □ u "' 0 Do u 0 105 0 ■ IO' -� 0.94 0.94 • □ • □ [ ■ IO' l0 ■ IO' 0 0.92 ■ D 3 0.92 □ 3 10 ■ 10 □ D 0.90 102 0.90 IO' 0 10 15 20 25 30 0 10 15 20 25 30

%Weathered (by mass) %Weathered (by mass) Figure 3-2. Density (circles) and viscosity (squares) at 0°C (solid) and 15°C (open) as a function of evaporative weathering forA WB (left), CLB (right).

From Figure 3-2, it can be seen that density increases with increasing evaporation were moderate. The density forthe AWB oil at 15°C ( open circles, left panel for Figure 3-2) increased from 0.9253 g/mL for the unweathered oil to 1.0140 g/mL to the most evaporated state of W4. In general, weathering increases density between 7�8% fromunevaporated to most weathered state. Temperature decreased oil density by approximately 2% from 0°C to 15°C.

In contrast to the relatively small changes in density, viscosity was found to increase by approximately a factor of 10 for each increasing stage of weathering. For example, for the A WB ° oil at 15 C (open squares, left panel for Figure 3-2), viscosity increased approximately 4 orders of magnitude over the total extent of the weathering tested, from 347 mPa•s forthe fresh product 6 to 7.91 x 10 mPa•s forthe highly weathered W4.

For the unweathered and comparably weathered fractions, the AWB diluted bitumen sample was measurably more dense and viscous than the CLB.

035 From Tables 3-1 and 3-2, for the A WB and CLB respectively, it can also be seen that pour point increased with increasing evaporation, from below -25°C to 27°C for AWB, for example. This paralleled the increases in viscosity, discussed above.

Flash point also increased with increasing weathering, reflecting the loss of volatile compounds in the materials by evaporation. Water in the oil was similarly lost by evaporation, with the more weathered fractions essentially free of water. As received, the CLB and AWB dilbit samples contained 0.6% and 1.5% water (w/w) respectively. However, this was lost during progressive evaporation stages. In contrast, total elemental sulphur did not appear to evaporate from the oil, and instead was increasingly concentrated in proportion to the mass lost by evaporation.

Surface and interfacial tensions of the more weathered products were not measurable, as it was not possible to form interfaces with controlled geometries for highly viscous materials. Measurements for the unweathered and lightly weathered Wl samples, however, show that the two diluted bitumen products had typical air/oil surface tensions (29 to 31 mN/m) and oil/salt water interfacial tensions (24 to 38 mN/m) for light to medium crudes in the Environment Canada database (Environment Canada, 2006).

As discussed in Chapter 2, similarities have been noted between diluted bitumen products and fuel oils (ENG, 2011). IFO 180 has a density of0.9664 g/mL at 15 °C (Environment Canada, 2006, reproduced in Table 2-2) and dynamic viscosity of 1920 mPa•s at 15 °C, which is comparable to the physical properties of "Wl" weathered versions of winter dilbits. For the AWB Wlsample the corresponding values were a density of 0.9531 g/mL and a viscosity of982 mPa•s (Table 3-1) For CLB, a density of 0.9537 g/mL and a viscosity of1330 mPa•s (Table 3-2) were recorded. To examine this comparison in more detail, reference has been made to IFO 180 throughout the present report. Comparisons are made with respect to the weathering behaviour, discussed below, and in the buoyancy fate discussions in Chapter 4. In the wave tank studies discussed in Chapter 5, weathered dilbit products similar to "Wl" in Chapter 3 are used. It is not the view of the authors of the present report that IFO 180 is a match or surrogate for diluted bitumen, but that comparisons between the properties, composition and behaviours of the diluted bitumen and IFO 180 may provide some insight on how spills of dilbit compare to conventional fuels.

One interesting difference between the IFO and the dilbits is variation of physical properties with temperature. Comparing the dilbit results to those of IFO 180, the temperature dependence of the viscosity for the fuel oil was significantly larger than that of the dilbits such that at the lowest temperature, the rate of change was found to be 5 to 10 times greater for the IFO 180 than the weathered dilbits.

The highly weathered versions (W2, W3 and W 4) of the winter dilbits may be more comparable to heavier intermediate fuel types or even a heavy fuel, such as HFO 6303, as shown 32

036 in Table 2-2. At 15°C, HFO 6303 has a density of 0.9888 g/mL and a dynamic viscosity of 241,000 mPa•s. Comparing these values to those of the weathered diluted bitumen in Tables 3-1 and 3-2, the density and viscosity of heavy fuel oil 6303 (Environment Canada, 2006, reproduced in Table 2-2) falls between the moderate W2 weathering state and the more highly evaporated W3 for both dilbit types.

Two types of oil transformation behaviour were tested for both products: evaporation and emulsification. The details of these tests are examined below. The effects of these processes on the density and buoyancy of the diluted bitumen are the subject of section 4, below.

3.4.2 Pan Evaporation of Diluted Bitumens

25 I 20 15 A\)/B CLB IFO-180 -- %Ev=2.398 ln(t) -- %Ev=2.187ln(t) -- %Ev=0.0455sqrt(t) 10 2 2 2 & r =0.928 r =0.993 r =0.992

0 200 400 600 800 1000 0 200 400 600 800 0 200 400 Time (hours) Time (hours) Time (hours) Figure 3-3. Pan evaporation results for two dilbit oils and IFO 180. Measured mass loss in an open pan as a function of time is shown as points, for AWB on the left figure, CLB in the central figure and IFO 180 on the right figure, respectively. Fitted simple expressions are shown as solid lines in red for each result. The results of the fits are indicated in the legends on each graph.

The evaporation rates for the two products were assessed by measuring mass loss from an open pan at 15° C, with air disturbance kept as low as possible, following the procedure described in Fingas (Fingas, 2006). The evaporation curves for the two crudes are shown in Figure 3-3. Test lengths for pan evaporation are typically 200 hours; however, to better determine the rate and final extent of evaporation possible, the time for the A WB was extended to almost 1000 hours, and over 600 hours for the CLB. The relative percentage of mass lost as a function of time (in hours) for AWB is shown on the left panel of Figure 3-3 and the middle panel for CLB, respectively. The evaporation curve measured for IFO 180 is also shown on the right-most panel, for comparison. Rates of evaporation for both the A WB and the CLB were found to be rapid compared to the IFO, as well as occurring to a much higher degree of mass loss. Within 6 hours at 15°C, the AWB had lost 15.9% of its mass, the CLB 11.7% of its mass. By comparison, the 33

037 IFO 180 had lost 0.85% of its mass. Within 24 hours, the relative mass losses were 18.4%, 15.6% and 1.75% for the AWB, CLB and IFO 180, respectively.

Both diluted bitumen products exceeded more than 20% mass loss: in 55 hours for AWB and 147 hours forCLB. The IFO 180 was not found to exceed 5% mass loss in a test of usual length: 200 hours. The weathering curve for HFO 6303 is not shown, as it has not been found to evaporate to more than 2.5% (Environment Canada, 2006).

Simple logarithms were fitted to the pan evaporation curves for both dilbits. The results are plotted on Figure 3-2. These curve fits were chosen based on the forms proposed by Fingas (Fingas 2004): %Ev = (A-0.045T) ln(t). For AWB, the best fit result was:

%Ev = (l.72-0.045T) ln (t)

forCWB:

%Ev = (l.51-0.045T) ln (t) where %Ev is the relative mass loss of the oil over time, as a percentage, Tis the temperature in degrees Celsius, and tis the time in minutes. Upon comparison with approximately 200 similar fit results for oils and fuels (Fingas, 2004), the factors of 1.72 for AWB and 1.51 forCWB indicate that both diluted bitumen samplesevaporate at an overall rate similar to a mid-weight (API 25-35) crude oil. However, it should be noted that for both cases, the simple logarithm models undershoot the early evaporation period of the measured curves (prior to approximately 50 hours) and overshoot the rate at the longer times. This is most noticeable for the A WB diluted bitumen on the left-most panel of Figure 3-3, but can also be seen in the CLB curve in the middle panel of Fi gure 3-3. While the overall rate may be equivalent to a crude oil, the shape of the logarithmic curve seems to underestimate the rate of evaporation early on, while overestimating later. Simple fit models may not be very useful for diluted bitumen, particularly in the early stages of a spill.

3.4.3 Simulated Distillation of Diluted Bitumens

The potential for evaporation of the products 1s also very commonly assessed by examining the distillation curves of the products. A common method for measuring distillation rates, temperature as a function of the fraction of oil distilled, is simulated distillation by gas chromatography, discussed in section 3.2.

The graphs in Figure 3-4 show simulated boiling point distributions ofCLB and A WB dilbit samples. On the left side of Figure 3-4, the boiling point distributions for fresh and moderately weathered (W2) fractions of the two dilbits are plotted. The graph on the right side of

Fi gure 3-4 shows the measured boiling point distributions for the two unweathered dilbit samples, a conventional crude (ASMB #5), and an intermediate fuel oil (IFO 180).

038 The low initial boiling points (the y-axis intercept on the left panel of Figure 3-4, at zero mass lost) for the fresh dilbits are evidence of the presence of low-molecular weight condensate in the dilbit blends. The elevation of initial boiling point in the weathered W2 fractions (seen as the vertical shifts between the fresh and W2 samples at low mass on the left panel of Figure 3-4) indicate that some of the "light-end," low mass molecules were evaporated from the W2 samples. However, at higher temperatures, the lines for the distillations converge, indicating that weathering preferentially removed the low-boiling range compounds of the oil prior to the high boiling range compounds.

800 �------800 �------,, . 700 f------fi�':.....+-'

0------0------20 40 60 80 100 20 40 60 80 100

·100 �------ -100 �------ Weight% Off Weight% Off

-CLB ---•CLB(l5.74%W) -AWB ---•AWB(l6.89%W) -CLB -AWB ••••••· ASMB #5 ----IF0-180

Figure 3-4. High temperature simulated distillation plots for the two dilbits before and after weathering (left) and in comparison to a conventional crude (ASMB #5) and intermediate fuel oil (IFO 180).

While both AWB and CWB are roughly similar in distillation curves, differences are apparent on close examination of Figure 3-4. The AWB curve shows higher mass loss at lower temperatures than the CLB curve, but at just over 200°C, the red and blue curves cross for both the unweathered and weathered samples on the left panel of Figure 3-4. This may indicate a higher concentration of "light-end" condensate in AWB compared with CLB, to compensate for a less-volatile, and thus presumably denser, bitumen feedstock for the AWB compared with the Cold Lake source bitumen.

On the right side of Figure 3-4, the distillation curves of dilbit samples are compared with IFO 180 and a crude oil (ASMB #5). The low-temperature boiling ranges for the diluted bitumens are much more similar to the crude oil than IFO 180, indicating a substantial difference in light, volatile compounds between the dilbit, made by blending with condensate, and the fuel oil. By comparison of the left panel with the right, it is apparent that even the moderately weathered W2 samples retained significantvolatile compounds compared with the IFO 180. This difference contrasts with the similarities noted for the bulk properties of density and viscosity

039 above, where the more lightly weathered Wl fractions were comparable to the IFO 180. These differences indicate the caution required in making direct comparisons between dilbit and conventional oils or fuel products.

3.4.4 Tendency to form Water-in-Oil Emulsions

The emulsion formation behaviour of the dilbit compounds was also examined. As can be seen in Tables 3-1 and 3-2 respectively, in most cases both the A WE and the CLE and their synthetically weathered subsamples formed continuous phases of oil containing discrete droplets of (simulated) seawater when mixed with 3.3% NaCl brine. Except for the fresh CLE/seawater mixture, these were all found to fall within the entrained water class of mixtures. Water contents ranged from 7 to 46% for all dilbit/seawater mixtures. However, in the case of the fresh CLE, a true emulsion formed, as can be seen by the elevated water content of 72% and complex modulus in Table 3-2. Complex modulus represents resistance to flow, which includes both inelastic (viscosity) and elastic properties of a substance. In general, for most oils and entrained states, this increases with viscosity. However, as can be seen in Table 3-2, the complex modulus for the unweathered state was observed to be higher than that of the Wl weathered fraction, 90.5 Pa and 64.0 Pa respectively. This indicates an elastic resistance to flow in the fresh CLE/seawater mixture, typical of true emulsion formation. The most-weathered form of AWE also did not incorporate a significantamount of water in the test. It was likely too viscous to trap water in the mixing vessel.

In their tendency to form emulsions with seawater, the diluted bitumens are comparable to the behaviour of IFO 180 and the heavy fuel HFO 6303 (Environment Canada, 2006). As can be seen from Table 2-2, both IFO 180 and HFO 6303 similarly form entrained mixtures with seawater, with water contents of 42% and 58%, respectively. The dilbit/seawater mixtures formed fromthe lightly weathered fractions, Wl, were found to be similarin both water content and complex moduli to a IFO 180/seawater mixture. However, HFO 6303 formed a mixture with 58% water content, much higher than the highly weathered W3 dilbit samples ( of comparable density and viscosity, as discussed above), which formed mixtures with lower water contents, 6 to 10%.

3.5 Chemical Composition of Diluted Bitumens

Figure 3-5 summarizes the analysis of the SARA groups for the AWE (left panel) and CLE (centre panel) dilbits. Information on these four main groups of chemicals are important both to the oil production and refining industries, but also have influences over the behaviour of oil, including evaporation and emulsification, as well as implications on the effects of oil spills, including toxicity of the oil. Also included on the right-most panel of Figure 3-5 are data for IFO 180 and HFO 6303 taken from the Environment Canada database (Environment Canada, 2006; also summarized in Table 2-2).

040 From Figure 3-5, it can be seen that as the weathering of the products increases with increasing evaporation loss, the proportion of saturates and aromatics both decreased while the proportion of non-volatile resin and asphaltene groups increased for both dilbit samples. The higher relative proportions of asphaltene and resin groups in all the AWB measurements in comparison with the CLB oils, fresh and evaporated, indicate that the A WB bitumen feedstock was composed of higher levels of high molecular-weight groups than the CLB bitumen. This is consistent with observations noted earlier concerning the simulated distillation data shown in

Figure 3-4.

Significant differences between the overall composition between the two dilbits and the conventional fuels are also apparent from examination of Figure 3-5. The relative content of asphaltenes was measured to be much lower in the fuel samples in comparison with the diluted bitumen samples. In addition, the relative propo1tion of saturate compounds was higher in the fuels than in the dilbit samples, especially with increasing evaporation of the dilbits. As asphaltenes and resins play significant roles in oil colloid and droplet formation, this finding could imply differences in behaviour between conventional fuels and dilbits.

Access WesternBlend Cold Lake Blend Fuels

100

- Saturates 60 � Aromatics � Resins - Asphaltenes_ 40

'" "' �o \ "'�i o f:,", "'

Figure 3-5. Changes to AWB, CLB hydrocarbon group composition with evaporative weathering, compared to conventional fuels.

Figures 3-6, 3-7 and 3-8 show the distributions of the saturated petroleum hydrocarbons, the petrogenic biomarkers, and the PAHs, respectively, for the AWB dilbit and simulated weathered subsamples. Tabulated data and datasets forthe CLB are available in the supplemental data set accompanying this report, but were found to be generally similar to the observations noted below.

Figure 3-6 shows the chemical abundances of n-alkanes and characteristic isoprenoids. The concentrations of alkanes n-C9 to n-C14 fell slowly with increasing weathering between the 37

041 fresh and W2 sample, but markedly decreased in abundance between the W2 and W3 samples. Similar trends are observable in the CLB data set.

Figure 3-7 depicts the abundances of the biomarker terpanes and steranes in AWB from fresh to W4. High abundances of biomarkers were detected in all these samples, at 1852, 1968, 2141, 2233, 2312 µg/g oil for the fresh, Wl, W2, W3, and W4 samples respectively. The increasing relative abundances can be explained by concentration of the refractory biomarkersin the residual weathered oils by loss of the lighter compounds. Similar to the findings of Wang and co-workers (Wang et al., 2011), biomarkers were found in abundance in samples of both AWB, as seen in Figure 3-7, and also in the CLB dataset. In comparison, total biomarkers in IFO 180 are 383 µg/g oil and 255 µgig oil in HFO 6303 (Environment Canada, 2006), almost 10 times lower than those in the oil sands products.

042 O"� >zj Concentration oil) "'l IJQ-· (µg/g "' C � "'l "'l � A co N °' .t. co w A o:i N 0\ .t. co w A co N °' 0 0 0 0 0 0 0 0°' 0 0 0 0 0 0 0 0°' 0 0 0 0 ....� I.,;' 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 "'l O'\ n-C9 n-C9 n-C9 n-C9 n-C9 '[::: n-ClO n-ClO n-ClO n-ClO n-ClO n-C11 n-Cll n-Cll n-Cll n-Cll �;:;·;;,. n-C12 n-C12 n-C12 n-C12 n-C12 - n-C13 n-C13 n-C13 n-C13 n-C13 �� n-C14 n-C14 n-C14 n-C14 n-C14 n-C15 n-C15 n-C15 n-ClS n-C15 "' :: n-C16 n-C16 n-Cl6 n-Cl6 n-C16 s- � n-C17 n-C17 n-C17 n-C17 n-C17 :: C. Pristane Pristane Pristane Pristane Pristane C. -· n-C18 n-C18 n-C18 n-Cl8 n-C18 Phytane Fhytane Phytane Phytane Phytane ::l � n-C19 n-C19 n-C19 n-C19 n-C19 C. O"-· n-C20 n-C20 n-C20 n-C20 n-C20 C. C n-C21 n-C21 n-C21 n-C21 n-C21 � n-C22 n-C22 n-C22 n-C22 n-C22 < -....· n-C23 n-C23 n-C23 n-C23 n-C23 n-C24 n-C24 n-C24 n-C24 n-C24 Q;;· g Ill n-C25 n-C25 n-C25 n-C25 n-C25 n-C26 n-C26 n-C26 n-C26 n-C26 0 o' n-C27 n-C27 n-C27 n-C27 n-C27 "'l n-C28 n-C28 n-C28 n-C28 n-C28 6; n-C29 n-C29 n-C29 n-C29 n-C29 n-C30 n-C30 n-C30 n-C30 n-C30 n-C31 n-C31 n-C31 n-C31 n-C31 n-C32 � n-C32 l;.) n-C32 � n-C32 n-C32 n-C33 n-C33 n-C33 ...... n-C33 § n-C33 n-C34 N n-C34 N n-C34 n-C34 n-C34 t"l n-C35 0 n-C35 Tj "'l n-C36 .I>, n-C36 l;.) n-C36 00 n-C36 v.. n-C36 "'l n-C37 v.. n-C37 .I>, n-C37 0\ n-C37 l;.) n-C37 al 0 n-C38 n-C38 n-C38 n-C38 n-C38 "'l n-C39 'J:- n-C39 'J:- n-C39 'J:- n-C39 'J:- n-C39

;'" O" >zj ti? ., "'l-· IJQ Concentration oil) -0� "' 5 (µgig � N \,> ... � N \,> ... I-' N !,,J A � N \,> ... � N \,> 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0... j; � 0 0 0 t 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 � "'l .... • � , ':> :r-s· 0:l ,t':> ,1') 1, ,') ,1,') ,t � � � ':> � ':> � ':> Q.-·,., 0 11,') 1, ,t 1t 1,t = -· � , � t':> �1 � , � 1,':> 1,':> 1,') 1') 1,') 12 3 �� , ;'" � .1, � ') �·1,')') �·1,')':> �·1,') �·1,')':> 1,') 1, 1,t 1,t "'�* �\, � \,1, � � � -�\ ':> � ':> ') - t - 1,':> - t ') ::- :: "'l 1, � 1, 1, 1,':> \, -�� 'c s;, s;, s;, ') s;,� 1 0 C. C. °'� � ,,1 ,') ') .,,o\, o.•"') • , ') ,') -0 � � '¢-o. o�') o'c"� �i:,1� -· <¢. ') <¢-".� -".� <¢-<>!'� ti? ., � o_-e\)r:,., o.•" <¢ �') ') ;: C. :!. <¢-lS''�,:, <¢ S''�,:, <¢- '' <¢-�,,�') <¢ S''� -l lS -l f" C. O" <¢- �,�') <¢- .-,,�') <¢- �, �') <¢-&'�') <¢- �,o ') ') 1, ') 1, ') <¢-t; � <¢-�1 ,� <¢ �1 ,� ') <¢- l,<,1 � <¢-�1, � � • a-• �, l� ') lS' lS' lS' - ') - -s,t�') 1,� ') � 0 "tilt� <¢-&1,� <¢- <¢-& ') <¢-&1,� �,:, ,, t ') 6; <¢- lS1� <¢-lS1��,:, '' lS1�� <¢-�,��') <¢-�,��') S �') �') ') ') <¢-&� <¢-&� <¢-&�� -&�� o :: ""' ') � ,:, � ,:, � <¢ ,:, <¢-& ,:, � <¢-lS'"� - S'"� - S1"� <¢-�,•� � <¢- S1"� Sl .I>. <¢ l W <¢ l N l - ii•� <¢- •�,:, - •�,:, -&•�') -&•�') <¢ l- � N (:il N <¢ (:il ...... <¢ ...... <¢ ,-3 > � "�,:, s ,:, 00 �,�') -• :;j <¢-lS1' (J'\ <¢-lS1'�') V.. <¢-lS1 0 <¢- \'o1 � <¢- ') ') ') ') ,,lS ') S� S� S� S� v, S, 0:l <¢- (:il t;<¢- (:il '.t: <¢-(:il g; <¢- (:il l;.) "&'� e1') <¢-��·,,p <¢-��•·P <¢- 1 <¢-��•'p �-. ��, \,, 1,, 'J:_ ') 'J:- ') �0 �·\, ,') 'J:- ') � � ') \-e

043 Figure 3-8 depicts the distributions of four series of alkylated PAH homologs in AWB and the evaporated samples: naphthalene (N), phenanthrene (P), fluorine (F) andch rysene (C), as well as the sulphur-containing series, dibenzothiophene (DBT). Degree of alkylation is indicated by the prefix, C#, where # indicates the number of methyl carbons. For example, C2-N indicates all the dimethyl and ethyl isomers for the naphthalene parent PAH. The sums of all measured isomers are shown as the vertical bars in Figure 3-8.

The relative abundances of the EPA priority PAHs are shown in a graph inset in the upper right corner of each respective alkyl P AH graph. In order of increasing molecular mass, these are biphenyl (Bph), Acenaphthylene (Acl), Acenaphthene (Ace), Anthracene (An), Fluoranthene (Fl), Pyrene (Py), Benz(a)anthracene (BaA), Benzo(b)fluoranthene (BbF),

Benzo(k)fluoranthene, (BkF), Benzo(e)pyrene (BeP), Benzo(a)pyrene, (BaP), Perylene (Pe), Indeno(l,2,3-cd)pyrene, (IP), Dibenzo(ah)anthracene(DA) and Benzo(ghi)perylene (BgP).

The total target PAHs in the AWB were 1919, 2140, 2331, 2265, 2246 µg/g oil for the fresh, 8.53% (Wl), 16.86% (W2), 25.34% (W3), and 26.45% (W4) evaporated fractions respectively. The fresh through W2 samples were both found to have characteristic bell-shaped distribution patterns for the naphthalene alkyl-PAR series, most likely from the lighter diluent component of the product. It can be seen, from Figure 3-8, however, that the phenanthrene, dibenzothiophene, fluorene and chrysene families all have a pattern of increasing abundance with increasing alkylation, typical of biodegraded oil, such as bitumen. For a single homologous family, the abundance of each alkyl derivative groupincreased with alkylation. For the fluorenes, for example, the relative abundances of the unalkylated fluorene: Cl-fluorenes : C2-fluorenes : C3-fluorenes were 22.8:75.8:185:251, respectively (expressed as µg/g oil). For the W3 and W4 samples, the naphthalene family also showed this pattern, indicating that the majority of the diluent components had evaporated. The alkyl PAH data forCLB were also very similar to this finding. This observation is again consistent with the findings of Wang and co-workers (Wang et al., 2011 ), who noted similar strong evidence of biodegradation in the patternsof the alkyl P AH that distributions they examined for both dilbit and native bitumen samples.

Differences were found again in the fuels in comparison to the diluted bitumen samples. The total alkyl PAHs in HFO 6303 was 28 300 µgig oil, while in the IFO 180, the total PAHs was 27 700 µg/g oil (Environment Canada, 2006), more than ten times higher than that found in any of the dilbit samples.

044 045 4.0 Oil Weathering and Buoyancy in Marine Conditions

4.1 Introduction

Following a spill, oil undergoes transformations in the environment. These processes, referred to as weathering, alter the physical properties and chemical composition of the oil. These chemical and physical changes drive the behaviour and fateof the oil in the environment

Oil density is the crucial factor in sinking. Studies have shown that oils with a density greater than 0.90 g/mL, and especially when this density approaches that of seawater (1.025 to 1.033 g/mL), can be over-washed (submerged under the water surface, but not sinking to the sea floor) under some sea conditions (Clark et al., 1987). The major mechanisms for oil submergence in water that have been observed include (NRC, 1999):

1. Evaporation increases an oil density with longer exposure (Wang et al, 2004). Evaporation alone, however, has been shown to be unlikely to increase density enough to cause sinking in seawater (Lee et al., 1989).

2. Photo-oxidation. Oils can oxidize in sunlight to form a dense crust on the surface of exposed oil (Bobra and Tennyson, 1989).

3. Water uptake. Mixing with water will not cause an oil to sink, but emulsion formation and/or water incorporation due to kinetic factors may change the properties of the oil such that sinking by other processes is more likely.

4. Mixing with solids can increase the density of the resulting mixtures. When mixed with silica sand, 2 to 3% mineral content in the oil phase is sufficient to cause sinking (NRC, 1999). Many studies have shown (see Chapter 2) that oil-sediment interaction is a common phenomenon.

Given the unknowns surrounding the uncertainty of the buoyancy of diluted bitumen products and the effects of environmental weathering and other factors such as mixing with seawater and sediment, a study was developed to investigate the range of possible behaviours of oil when subjected to these four types of weathering.

The objective of this study, reported in summary here, was to observe the range of possible fates of oil in salt water, when subject to evaporation, photo-oxidation, water uptake or mixing and sediment uptake or mixing. A series of experiments was designed to test the possible range of behaviours that might be exhibited by dilbits in the marine environment.

046 The following combinations were examined:

• Oil types: AWB, CLB, IFO 180 • Evaporation: Fresh, Wl, W2, W3, W4 (note that only fresh was evaluated for IFO 180) • Photo-oxidized: Phox-W2 • Salt water brine + oil mixing: Fresh through W 4 (only fresh for IFO) • Salt water brine + photo-oxidized oil: Phox-W2 • Salt water brine + oil + kaolin clay: Fresh throughW 4 (only freshfor IFO) • Salt water brine + photo-oxidized oil + kaolin clay: Phox-W2 • Salt water brine + oil + sand: Fresh, W2 • Salt water brine + oil + diatomaceous earth: Fresh, W2

The intent of the study was not to examine mechanisms of formation, but the possible end states of the processes, i.e., their fates. Each of the factors would be tested at the limits of what would be possible in the natural environment: high mixing energy conditions, high sediment loads, long evaporative and photo-oxidizing exposures. These conditions are discussed in the following section.

4.2 Materials and Methods

As this study was designed to examine the extent of behaviours possible, it was desirable to use high but still realistic influences for each of the factors. The production of samples for each of these factors is discussed in detail below.

4.2.1 Oil Types

The two types of diluted bitumen, CLB and A WB, were studied for their weathering behaviour, as well as resulting chemical and physical properties (see Chapter 3), to determine the relationship of oil properties to fate and behaviour for applications such as spill modelling and emergency response tactics. An intermediate fuel oil used for marine ship fuel, IFO 180, was also included for comparison.

4.2.2 Evaporation

Oils were evaporated m a rotary evaporator as described in Chapter 3, with known pressure and temperature to simulate evaporative weathering. As described in Chapter 3, this method of evaporation has been found to result in oil chemically similar to oil evaporated under ambient conditions from an open pan (Fieldhouse et al., 2010). Four fractions of weathered samples were generated, yielding different degrees of weathered fractions: fresh sample, 1/3 and 2/3 weathered (by mass lost), and finally 48 and 96 hours of simulated weathering, representing the end condition in the environment, designated as Wl, W2, W3 and W4 respectively. 43

047 The intermediate evaporation fractions (W2: 16.9% for AWB and 15.8% for CLB, respectively) of both dilbit samples were singled out for more comprehensive testing. The regulations for response organizations mandate a maximum response time for oil spill clean-up. Transport Canada requires that, for marine spills, all clean-up operations be complete by 10 days after the spill (Transport Canada, 1995). Based on the pan evaporation experiments, described in Chapter 3, and the rate of weathering reported by SL Ross (SLRoss, 2013), the W2 fractions were the closest match to a 240-hour exposure.

4.2.3 Photo-oxidation

No standard or widely-used methods exist for photo-oxidation of oil or petroleum products. Some authors have previously reported pilot studies of photo-oxidation of hydrocarbon mixtures, but no widely agreed-upon method exists in the literature.

Photo-oxidation was conducted in a controlled UV Crosslinker (SpectroLinker Inc.), which was calibrated to deliver a known energy flux at a wavelength of 315 nm. The incident energy flux of the photo-oxidizer was used to calculate an exposure time to simulate 120 hours of daylight exposure (10 day cycles). This was determined by integrating the ASTM G 173 solar referencespe ctra (version 2.9.2), to determine equivalent number of ionizing photons at 315 nm. This number of total photons for exposure was then used to estimate an exposure interval at the instrument's incident flux. Temperature was monitored to ensure that evaporative exposure during the oxidation period was minimized. The 120-hour exposure time was determined as the maximum likely exposure during the 10-day maximum clean-up period allowed under Canadian regulation (Transport Canada,1995).

One advantage of the closed system method described above is that both the photon and ionizing energy flux applied to the oil can be precisely quantified. Geometry is well defined, and light canbe applied in short "doses" to minimize the effects of heating. Other researchers have used ultraviolet sources with known flux (SL Ross, 2013), but because of limitations of the experimental apparatus, they did not quantify the effect of light exposure heating on the rate of evaporation of the oil.

The exposure of the oil was conducted under the narrow frequency range (peak315 nm) of the UV lamps used. The exposure spectrum was not a true solar spectrum andcould possibly cause differences in both the species generated by natural solar exposure. Calibration was also done to match the ionizing p01tion of the ASTM G173 spectrum. The limitations of that spectrum, calibrated for the purposes of solar panel testing, will be reflected in the length of ionizing exposure. Further validation of this procedure is planned.

4.2.4 Oil-Water Mixtures

Oil-water mixtures were prepared using a rotary end-over-end mixing of a 1: 10 ratio of oil-to-water at a rate of several revolutions per minute for up to 12 hours at 15 °C. This procedure 44

048 has been shown to provide a high-energy mixing environment, with sufficient mixing time to ensure that if a mixed water-oil state is possible, it will likely be reached in the test. In the case of high-viscosity heavy oils, the dominant form of result observed is the "entrained-water" state (Fingas and Fieldhouse, 2006), which has a typical water content of 30%. Chemically-stabilized stable or meso-stable emulsions of intermediate and heavy fuel oils are uncommon.

4.2.5 Oil-Water-Sediment Mixtures

In the event of a spill into water, both suspended and bottom sediments can interact with oil, impacting its fate. Interactions oftenresult in the formationof oil-sediment aggregates, which may become denser than water and sink. Oil-sediment aggregates that enter the water column may also release oil droplets over time (Lee et al., 2002). While natural sediments contain a range of sizes from sub-micrometre to millimetres, size-segregated sediments may give insights into sediment interactions as a function of particle size, which might lead to different fates (Aijiolaiya et al., 2006). Natural sediments may also contain a significant amount of organic material, which can affect oil-particle interactions as well. To simplify interpretation for the present work, mineral-only sediments were used.

Three size ranges of particles were selected. Different sediment size ranges may be encountered by oil, by suspended paiticulates in the open ocean, from a river mouth bearing sediment, and by contact with beach and shore sediments. This guided the selection of the sediments used in the present study. Thesediments used include:

1. Kaolin - hydrated aluminum silicate, Al2Si2O5(OH)4. This fine particulate typically falls in the range of 1-2 micrometre in diameter. It is acidic when mixed with water. Brine mixtures were neutralized with sodium hydroxide prior to use. Kaolin was also selected to allow for comparability with the sediment addition experiments reported in Chapter 5. 2. Diatomaceous earth - silicon oxide, SiO2. The medium-sized particulate had an approximate mesh size of 325, about 44 micrometre diameter. 3. Sand - silicon oxide, SiO2• Approximate mesh size range 50-70, or 200-300 micrometres in diameter.

The sediment loading chosen was 10 mg sediment/L brine. While high, this level of suspended sediment has been found in coastal river outflows. In a set of 44 estuaries measured in the deltas at the site of maximum turbidity at high tides, depth-averaged suspended sediment concentrations were reported from 1.0 to 48 400 mg/L (Uncles et al., 2002). For the Fraser River delta in British Columbia, surface suspended sediment concentrations have been reported approaching 1 g/L during the highest annual flows (Kostaschuk et al., 1993). Amos (Amos, 1996) reported a maximum depth-averaged suspended particulate matterconcentration of 3.5 g/L in the Bay of Fundy.

049 Oil-sediment interaction was assessed by first preparing a sediment-water mixture of 10 mg/L sediment in 600 mL of 33 ppt NaCl brine in a 2.2-L vessel. The water pH was checked, then adjusted to 7.0 when necessary, and allowed to equilibrate.

A 30-mL portion of oil was then added to the vessels. The mixing bottles were then sealed and allowed to thermally equilibrate for at least 4 hours. The vessels were mixed for 12 hours on the rotary end-over-end mixer at 15 °C. If the sediment interaction resulted in a well dispersed mixture with no separate phase, the contents were poured into 1-L graduated cylinders to monitor settling. If water-in-oil emulsions were formed, the emulsion phases were collected and transferred to wide-mouthed glass jars for observation and analysis (water content, viscoelastic properties, just after mixing and at 1 week).

4.3 Results

4.3.1 Evaporative Weathering

1.10 1.10 Access Western Blend Cold Lake Blend Conventional Oil and Fuels 1.05 1.05 Saltwater Saltwater Saltwater • ■ W3&4 � !W3&4 HFO 6303 1.00 Freshwater Freshwater • Freshwater 1.00 ;3' - .. W2.... - ---_,, 6 ·"----=-=.:::.::::··· "' -- -- ···0--- � - .. 0 - t,,,... -�i --- -- IFO 180 . 1:,,.___ _f'. . . 0.95 OW! 0 Wl ..• __ ··--·-f'.. 0.95 ·;: - ·- • - - - 0.90 • • • - ...... 0.90 AWB Fresh ..... CLB Fresh

Alberta Sweet Mixed Blend 0.85 ...._ 0.85 ...... 0.80 0.80 10 20 30 40 0 10 20 30 40 0 10 20 30 40

Temperature {°C) Temperature {°C) Temp era ture ('C) Figure 4-1. Effects of evaporative weathering changes on density as a function of temperature for AWB and CLB samples. Data on conventional fuelsand a light crude type are provided for comparison. Lines for fresh water and natural seawater from Spieweck and Bettin, 1992 and UNESCO, 1981, respectively.

The effects of evaporative weathering are plotted on Figure 4-1 for A WB on the left panel, CLB in the middle panel, and conventional oils and fuels for comparison on the right panel. Experimental measurements are shown as individual points. Note that standard errors, determined from the standard deviation of triplicate measurements, are smaller than the points plotted on Figure 4-1, and thus are not shown. Linear regressions fitted the curves of the fresh and moderately evaporated samples ("W2"), shown as lines. Details of the regressions can be found in Table 4-1. Lines for seawater (UNESCO, 1981) and fresh (distilled) water (Spieweck and Bettin, 1992) are shown for visual reference.

050 From Figure 4-1, it can be seen that the density of the fresh A WB and CLB products, approximately 0.92 g/mL at 15°C, rises to approximately 1.01 to 1.02 g/mL at 15°C at the maximum level of evaporation achieved in the W 4 samples.

Table 4-1. Linear regressions for densityas a function of temperature formeasured oils. 2 Density (g/mL) vs. Temperature {°C) Linear Fit r AWB Fresh p = 0.937 - (0.000847 * T) 0.935 AWB W2 p = 1.005 - (0.000870 * T) 0.999 CLB Fresh p = 0.937 - (0.000660 * T) 0.955 CLB W2 p = 0.993 - (0.000650 * T) 0.999 HFO 6303 p = 1.001 - (0.000732 * T) 0.994 ASMB#5 p = 0.846 - (0.000460 * T) 0.999 IFO 180 p = 0.981 - (0.000860 * T) 0.999

As canbe seen from Figure 4-1, the change in the density for both oil sands products was found to be linear with temperature. This is consistent with the usual behaviour for natural petroleum and refined products, as canbe seen by comparison with heavy and intermediate fuels oils (HFO 6303 and IFO 180, respectively) and a crude oil (ASMB #5). As described previously in Chapter 3, the density line of the IFO 180 is most similar to the Wl evaporated samples of both types of dilbit. The HFO 6303 data falls between the W2 and W3 evaporated dilbits, in both cases. The density of reference conventional oil, Alberta Sweet Mix Blend #5 (ASMB #5) was considerably lower than that of the dilbits (Figure 3-2) for all temperatures.

These results indicate that evaporation in combination with typical marine temperature ranges seen in Canada (0-l 5°C) is not sufficient to cause oil sinking for both the A WB and CLB products in fully salt marine waters, even at temperatures approaching freezing.

4.3.2 Photo-oxidative Weathering

Photo-oxidation for the simulated 120-hour exposure (10 xl2-hour days) produced a visible change in the oil from the unexposed weathered sample. Photo-oxidation of oils resulted in surface temperature change, and often with time pH decrease in the water phase, from pH 7.0 initially to pH 6.3 to 6.8 for the CLB and pH 6.19 to 6.72 forthe AWB following light exposure. There was crust and film formation on the oil slick, as can be seen in Figure 4-2. However, the photo-oxidized samples for both CLB and for A WB did not have densities significantly different from the corresponding non-UV-exposed W2 samples.

051 F1gure 4-2 Cold Lake dilbit (W2, 15. 75% evaporated) slick on 33 g/L NaCl water, post-photo-oxidation. There is distinct crust formation at slick edges. Indiscernible in the image is the presenceof a transparent film on oil surface, which formed after photo-oxidation. Photo: B.P. Hollebone, Environment Canada

4.3.3 Mixing with Salt Water

The results of the oil-saltwater mixing trials are discussed in detail in Chapter 3, and presented as a summary in Figure 4-3, for AWB. Results for CLB were similar to the example of AWB shown in thefigure. As described in Chapter 3, the results of mixing all evaporative stages of AWB with saltwater were all classified as entrained-water-oil mixtures. Water content was lower than the true emulsified states, ranging from 6% to 46%. As can be seen from Figure 4-3, while water uptake increased the density of the resulting mixture, in all cases the resulting mixtures remained buoyant in seawater(UNESCO, 1981).

Comparing the density increases upon mixing between the oil-saltwater mixtures for un exposed W2 sample (W2 for the oil and W2 Mix for the oil-saltwater mixture, respectively) and the photo-oxidized W2 samples (indicated on Figure 4-3 as W2-Phox and W2-Phox Mix for the photo-oxidized W2 AWB oil and the UV-exposed W2 AWB oil-saltwater mixture, respectively), a small increase in density can be seen for the photo-oxidized mixture over that for the unexposed sample. The densityof the W2-Phox mixture was approximately 1 % greater than that of the unexposed W2 mixtures. However, water content was found to be very similar for the two AWB mixtures: 33% for the W2 and 32% for the W2-Phox samples, respectively.

For CLB, however, mixture water content with and without UV exposure increased from 38% for the W2 to 48% for the W2-Phox saltwater mixtures. The density of the two CLB

052 mixtures, however, was approximately only 1% different, 0.9887 g/mL and 0.9767 g/mL for the CLB W2 and CLB W2-Phox oil/saltwater mixture densities respectively.

The density of the W3 oil-saltwater mixture appeared to be lower than that of the starting oil. Water content of the W3 mixture increased slightly from the starting oil, from 0% to 6% (refer to Table 3-1 ). The reason for this depression in density for the AWB W3 sample is not well understood, but was verified in replicate testing for this sample.

Mixture results for the most highly weathered AWB W 4 sample are not shown as no water uptake was found for this sample; no mixture formed. Also, as previously discussed in Chapter 3, a similarly-buoyant entrained-water state formed during the IFO 180 mixture test as well.

1.04 Access WesternBlend Seawater (15°C) 1.02

1.00 Freshwater (15°C) 0.98 � _g 0.96

0.94 I a: 0.92

Figure 4-3. Densities of AWB samples before and after mixing with 33g/L NaCl saltwater at 15°C. Arrows indicate the relationship between oils before and after mixing. Reference lines indicate the density of fresh and natural seawater at 15°C. The last point (W3 Tarballs) is described in section 4.3.4.

053 4.3.4 Mixing with Salt Water and Sediment

The results of oil/saltwater/sediment mixing tests are shown in Table 4-2. In general, mixtures with kaolin formed finely divided oil-sediment particles that sank in the salt water, with one exception for the highly weathered fractions for both AWB and CLB. Mixtures with the larger particles of diatomaceous earth likewise dispersed and sank, though a portion of the oil remained unmixed and floating on the surface. Mixtures with the coarser sand, however, formed a single, continuous phase of floating oil-saltwater entrained mixtures, and only a small portion sank as an oily sediment mixture.

Table 4-2. Physical fateand buoyancy of oil-saltwater and oil-saltwater-sediment mixtures. No Sediment Kaolin Diatomaceous Earth Sand (1-2 µm) (44 µm) (200--300 µm) AWB Fresh Emulsion - Entrained Disperse Disperse Emulsion-Entrained Float Sink Sink Wl Emulsion - Entrained Disperse Float Sink W2 Emulsion - Entrained Disperse Disperse Emulsion-Entrained Float Sink Sink W3 Emulsion - Entrained Tarballs Float Float W4 No change W2- Emulsion - Entrained Disperse Phox Float Sink CLB Fresh Emulsion-Mesa Disperse Disperse Emulsion-Entrained Sink Sink Wl Emulsion - Entrained Disperse Float Sink W2 Emulsion - Entrained Disperse Disperse Emulsion-Entrained Float Sink Sink W3 Emulsion - Entrained Tarballs Float Float W4 Emulsion - Entrained Tarballs Float Float W2- Emulsion - Entrained Disperse Phox Float Sink IFO 180 Fresh Emulsion - Entrained Disperse Disperse Emulsion-Entrained Float Sink Sink

Kaolin, when mixed with the oils, had a dispersive effect. When mixtures were poured into columns to settle, several phenomena were observed (see Figure 4-4). The fresh oil and three of the evaporated subsamples, as well as the photo-oxidized intermediate subsample (fresh, Wl, W2 and photo-oxidized W2 AWB and fresh, Wl, W2 and photo-oxidized W2 CLK), formed dispersions that settled to the bottom of the saltwater columns with time. With an increasing level of evaporation, the settled layer of oil-sediment became more compacted (see Figure 4-4). However, in contrast to the less-evaporated fractions, the highly-evaporated third and fourthfractions formed a state similar to tarballs which remained floating (see Figure 4-5).

054 When measured, these floating tarballs were found to have a density of 1.006 to 1.009 for the AWB W3 and 1.004 to 1.005 g/mL for the CLB W3 sample. For AWB, the density of the tarballs (W3 Tarballs) formed compared with those of the fresh oil (W3) the oil/saltwater mixture (W3 Mix).

Figure 4-4. Mixtures of Cold Lake dilbit (2013) and kaolin sediment (10 g/L) in salt water (33 g/L NaCl). In each image, fromleft to right: 0%, 7.88%, 15.75% weathered. Sedimentation appears more prevalent for less weathered oils: (a) after 1 hour, (b) after 3 hours, (c) after 24 hours, (d) after 1 week. Photo: B.P. Hollebone, Environment Canada

055 "' ,,,- •

Figure 4-5. Mixture of the W3 fraction of AWB (25.34% mass loss) and 10 g/L kaolin sediment. Photo: B.P. Hollebone, Environment Canada

Diatomaceous earth (DE) interacted similarly with the oils to form a dispersed mixture as can be seen in Figure 4-6 for the fresh and moderately weathered 0N2) AWD, but which settled over time more quickly than kaolin. From examination of Figure 4-6, the appearance of the oil DE aggregates appear similar; however, significant amounts of oil also remain floating on the surface of the saltwater columns. Testing this floatinglayer showed that the oil was similar to the water-oil mixtures formed without sediment present. The volume of the settled phase was also significantlylower thanthat of kaolin samples, suggesting that the higher particle size may have resulted in a denser final state.

The typical result of sand mixing with diluted bitumen oil can be seen in Figure 4-7, for the example of fresh and moderately weathered (W2) CLB. Results for AWB were very similar. The majority of the oil remained floating on the surface of the saltwater column. Testing revealed that the oils were emulsified and were very similar to the mixtures formed in the oil/saltwater mixing tests without sediments, assessed in terms of water content and complex modulus. However, as can also be seen in the bottom of Figure 4-7, some of the oil was found to be coated in a thin layer of sand, and had sunk to the bottom of the saltwater column as larger blob-like particles. These blobs were observed to shed large droplets of oil periodically; the oil would then riseto rejoin the surface layer.

056 ' . ... -� ••

Figure 4-6. Mixture of the fresh (left) and W2 (right) fractions of AWB and 10 g/L diatomaceous earth sediment after 24 hours of settling. Note compact oil-sediment layer on bottom and oil layer on top. Photo: RP. Hollebone, Environment Canada

- • •

Figure 4-7. Mixture of the fresh (left) and W2 (right) fractions of CLB and 10 g/L sand sediment after 24 hours of settling. Photo: B.P. Holle bone, Environment Canada

057 While single examples of all mixing experiments are shown in Figures 4-5 through 4-7, substantially similar results were found for both diluted bitumen samples: (1) dispersal and sinking of finely divided particles when mixed with the kaolin for the fresh to W2 ( and W2- Phox) samples; (2) tarball formation when mixed with kaolin for the highly weathered W3 (and W4) samples; (3) dispersal and sinking of finely divided particles, with some oil-saltwater mixtures on the top of the column for the DE trials; and ( 4) very little mixing with silica sand, with the majority of the oil forming oil/saltwater mixtures on the surface of the column, with a few large sand-encrusted larger blobs that sank. Trials of IFO 180 mixed with saltwater alone and with the three sediment types showed over-all behaviour for this fuel to be similar to the moderately-weathered A WB and CLB W2 fractions.

4.4 Comparisons with Literature and Historical Spills

The results of the present study, when compared to the findings of SL Ross (SLRoss, 2013), agree with many of the SL Ross results for oil-saltwater mixing. Water uptake for fresh oil was observed to be higher for the fresh CLB (72% w/w) in the present work than in the SL Ross flume, 39% and 34% (v/v) in the two trials, but that may be explainable by the much lower mixing energy available in the SL Ross flume.

Understanding the differences in weathering methods, rotary evaporation for the present work, open exposure on-water with wind for the SL Ross studies, it would seem that the W2 fraction results in the present work are likely the most similar to the end-states in the SL Ross experiments. On this basis, the W2-saltwater mixtures forCLB were found to be 38% water and 48% for the W2-Phox-saltwater mixtures, while SL Ross found that water contents fell to 25% for the trial with no UV exposure and 20% with UV exposure after their tests were concluded in 120 and 300 hours, respectively. Again, the different mixing energies available between the tumbling flasks used in the present study and the overflow weir used in the SL Ross flume may be the best explanation of these differences. One significant finding from both studies is the uncertainty in the effect of photo-oxidation on water uptake during mixing: in the present work for the AWB and CLB photo-oxidation results (similar water uptake for AWB, more for CLB) and again between the CLB results found by SL Ross (lower water uptake with UV exposure).

The results of the present work also match qualitatively with observations made at the Burnaby 2007 spill of Albian Heavy Synthetic (TSB, 2007). In that spill, oil flowed out of a sewage discharge into Burrard Inlet in conditions of clear weather and moderate wind. No oil was reported or found to have sunk, based on sediment and water quality monitoring following the spill (Stantec Consulting Ltd., 2012b). Given the speed of the clean-up operations, and the calm conditions that prevailed, it is unlikely that any oil-water mixtures or oil-sediment mixtures would have been likely to form. As shown in Figure 4-1, under these conditions, the oil would be expected to float, as observed, even if extensively evaporated.

058 During the Kalamazoo spill of 2010, the EPA documents the spill of a mixture of two types of dilbit, A WB and Canada Western Select (EPA, 2013) into a tributary of the Kalamazoo River, flowing quickly into the river itself. The oil was carried 40 miles downstream during a flood. While much of the oil was removed from surface water and the shorelines following the spill, an estimated 10% to 20% of the oil mixed with sediment in the turbulent flood water and sank to the bottom of the river (EPA, 2013). Epifluorescence micrographs of the sunken oil particles (Lee et al., 2012), indicated small oil droplets encrusted with sediment. Qualitatively, the oil-sediment agglomerations in the Kalamazoo River (Lee et al., 2012) appear to be very similar to the structures observed in the kaolin and diatomaceous earthtests in the present work.

059 5.0 Wave Tank Studies: The Fate and Behaviour of Selected Diluted Bitumen Products in Natural Sea States and Environmental Conditions

5.1 Introduction

The objective of this phase of the research initiative was to evaluate the natural attenuation and the application of a conventional oil spill treating agent and mineral fines to determine how they influence the behaviour, fate and transport of diluted bitumen (dilbit) spilled at sea. Two synthetically weathered dilbit products, A WB and CLB were treated with a chemical dispersant, sediment (e.g., mineral fines) and a combination of the two and compared to natural attenuation (no treatment), in limited hydrodynamic and environmental conditions. All experiments were conducted in seawater at average temperatures of 8.3 ± l.3°C.

5.2 Materials and Methods

5.2.1 Wave Tank Studies

The wave tank facility where the workwas undertaken is located at the BedfordInstitute of Oceanography (BIO) in Dartmouth, Nova Scotia. Tank dimensions are 2 m high, 0.6 m wide and 32 m long. Further details on the wave tank can be found in Li et al. (2008). The wave tank is capable of producing breaking and non-breaking waves. As previously reported for breaking waves, the energy dissipation rate decreased from 1.0 x 10-2 watts/kg a few centimetres below the surface to 5.0 x 10-4 watts/kg 20 cm deep into the water column. Non-breaking waves 6 produced an energy dissipation rate of 5.0 x 10- watts/kg 20 cm deep into the water column (Wickley-Olsen et al., 2008). Calibration of wave energies is an ongoing exercise, as new and improved technologies associated with hydrodynamics are continuously introduced to the wave tank facility. The wave energy dissipation rates reported are similar to those measured in the field by Terray et al., 1996 and Drennan et al., 1996. Four water sampling devices were deployed, one at 2 m upstream from the oil release point and the other three at 2 m, 10 m and 14 m downstream from the oil release point. Each of the four samplers collected water (-100 mL) at three depths (5, 75 and 145 cm) from the tank at various time points (5, 15, 30, 45 and 60 minutes). In addition, surface and effluent samples were collected. Samples (t=0) were taken at location "D" (14 m downstream) during each experimental run. The factorial design in Table 5-1 was conducted in random order to reduce confounding effects from wind, rain, seawater salinity and temperature. The selected oil spill treating agent has been successfully used in the past to treat conventional crude oils under various conditions (Li et al., 2008; Li et al., 2009a&b; Li et al., 2010). 5.2.2 on Dispersantand Mineral Fines Application For each experimental run, oil was released from a 40 cm i.d. containment ring, following the procedure used in Li et al., 2008. Dispersant applied by pressurized sprayer to produce a 56

060 uniform distribution of dispersant onto the surface of the oil slick at a dispersant-to-oil ratio of ~1 :20 (Li et al., 2008; Li et al., 2009a&b; Li et al., 2010).

Mineral fines (kaolin, see description in Chapter 4) were prepared as a seawater slurry (1 :20 w:v). The slurry was applied by a hand-operated sprayer. This provided even distribution of fines over the oil surface and gentle application preventing minimal disruption of oil on the surface. As an additional treatment option, mineral fines were used in combination with dispersant application.

5.2.3 Wave Tank In-situ Measuring Devices

Two particle size counters (LISST-lO0X, Sequoia Scientific, Inc.) were employed during the experiments, at 1.2 m and 12 m downstream of the oil release point and at a depth of 45 cm. Particlesize distributions were recorded every 2 to 5 seconds for1 hour per experiment (Li et al., 2008; Li et al., 2009a; Li et al., 2010). Fluorometers (Turner Cyclops C7) were attached in-situ in the same locations as the LISST particle size analyzers to monitor changes in fluorescence produced from dispersed oil droplets entering the water column.

5.2.4 Lab Analysis of Seawater Samples

Subsamples (20 mL) of collected water samples were used to measure surface tension in a temperature-controlled environment (20.7 ± 0.4° C) following the Wilhelmy plate method (Drelich et al., 2002). A Hydro-Bios surface sampler was deployed to collect surface samples from the top to a depth of 1 cm (sea surface microlayer). Details on the design of the sampler are found in Guitart et al., (2008), and the sampler is commercially available through the supplier, Hydro-Bios. Water samples (surface and subsurface) were extracted and analyzed for total petroleum hydrocarbons (TPH) using a gas chromatograph equipped with flame ionization detection according to the method of Cole et al. (2007).

Extracts of seawater surface samples were evaluated for SARA compound groups using thin-layer chromatography coupled with flame ionization detection (Maki and Saski, 1997). Some of these extracts were further processed to measure PAHs and their alkylated homologues using a gas chromatograph interfacedto a mass spectrometer (EPA method 8270 C).

061 5.3 Results and Discussion

5.3.1 SyntheticallyWeathered Di/bit Products

Evaporation is expected to play an important role in oil weathering in the initial stages of a spill. Light oils lose up to 75% of their initial volume, medium oils lose 40% of their initial volume, and heavy or residual oils will lose ~5% of their volume in the first few days of an oil spill (Fingas, 1999). Synthetic weathering of the two oil products prior to testing in the BIO wave tank facility resulted in losses of >6% of their mass (weathering process cited in Chapter 3), which, based on Fingas, may represent a couple to a few days of weathering in the natural environment, a period of time that might be expected before spill response measures can be initiated in some incidents. Synthetically weathering oils removes volatiles, thus reducing experimental variance due to evaporation (Li et al., 2008).

5.3.2 Wave Tank Experiments

Initial wave tank studies of dilbit dispersion were carried out in a seawater environment (8.3 ± l.3° C) that represents surface water temperatures reported during springtime in Douglas Channel (Stronach et al., 2010). As outlined in Table 5-1, salinity ranged from 25 to 30 parts per-thousand (ppt). Salinities in Douglas Channel vary from less than 20 to 33 ppt depending on location and seawater depth (ASL Environmental Sciences, 2010). The humidity values were included, which provide an indication of precipitation such as rainfall. The introduction of freshwater can be a confounding factor of oil dispersion effectiveness. In these limited environmental conditions, most of the wave tank experiments were carriedout in breaking waves with current flow.

A limited number of treatments including natural attenuation and the addition of mineral fines anddispersant to the weathered products were assessed in non-breaking waves with current flow. These were conducted to illustrate that, as a minimum, breaking wave energy is required to maximize the effectiveness of oil spill treatments. Other oil recovery methods such as mechanical skimmers and booms are more effective during calm sea states. Figure 5-1 is a collage of photos from wave tank studies illustrating the behaviour of dilbit in the presence and absence of chemical dispersant in non-breaking and breaking waves.

062 Table 5-1. Wave tank experimental conditions. Each grouping represents replicate runs of the same conditions (AWB = Access Western Blend, CLB = Cold Lake Blend).

Corexit Kaolin Water Wind Air Oil Oil Humidity Salinity Exp.# Slurry Temp. Speed Temp. Type (g) 9500 (%) (ppt) (g) (g/L) {°C) (km/hr) {°C) 1-1 AWB 253.60 12.08 0 6.4 6.1 6 12.2 29.2 2-1 AWB 238.98 15.37 25 7.6 5.8 6 5.5 29.0 3-1 AWB 247.80 0 0 8.0 4.6 5 11.0 29.5 4-1 AWB 240.84 0 25 6.2 5.2 5 6.9 29.2 5-1 CLB 264.60 12.17 0 6.6 6.7 7 6.4 29.1 6-1 CLB 267.76 13.51 25 8.0 4.1 4 10.0 29.3 7-1 CLB 295.40 0 0 7.6 16.8 17 11.2 28.8 8-1 CLB 232.80 0 25 6.5 5.7 6 8.9 28.5

1-2 AWB 252.29 12.15 0 9.0 3.1 3 11.2 29.4 2-2 AWB 276.07 12.99 25 8.9 8.2 8 8.0 29.0 3-2 AWB 264.80 0 0 9.7 16.2 16 10.8 27.9 4-2 AWB 272.60 0 25 8.5 8.2 8 22.0 29.3 5-2 CLB 264.45 12.72 0 10.5 23.8 24 16.0 25.9 6-2 CLB 257.22 13.58 25 9.5 3.9 4 10.0 25.6 7-2 CLB 259.76 0 0 9.0 19.0 19 6.0 26.7 8-2 CLB 262.42 0 25 9.9 13.6 14 10.0 26.9

1-3 AWB 276.05 12.54 0 7.5 4.0 4 11.6 29.4 2-3 AWB 260.05 12.77 25 9.2 19.0 19 9.2 28.2 3-3 AWB 293.81 0 0 9.5 15.2 15 14.2 27.5 4-3 AWB 266.76 0 25 6.9 2.9 3 7.0 30.0 5-3 CLB 259.78 12.01 0 8.7 19.6 20 13.0 28.9 6-3 CLB 278.00 13.19 25 9.4 5.5 5 9.4 27.9 7-3 CLB 237.27 0 0 9.2 22.1 22 8.0 24.8 8-3 CLB 254.71 0 25 7.0 5.6 6 8.2 29.4 Avg 261.58 12.92 8.3 10.3 28.3 STDev 15.96 0.94 1.3 3.6 1.4

063 Figure 5-l. A collage ofpltotos: A) Wave tank facility; B)Non-breakingwaves; C) B,,,akingwaves; D) Dil>it as a non-dispersed slickin non-b ,,..kingwave (no diopersant); E) Dilbit coalescenceand surface slick formationin breakingwaves (no diopersant); F) Dil>it as a no n-diopersed oil slick inJtOn-breaking waves (witlt diopersant); and G) Chemicallydispersed dil>it in tlte subsurface water colwttnin breaking waves. Photo: TL. King, Dep artment ofFislteries and Oceans.

064 5.3.3 Surface Water Profiling

In breaking waves, the dilbit products with no oil spill treatments were driven into the subsurface water column; however the products resurfaced as large oil droplets (e.g., >75 µm) and coalesced at the surface, forming an oil slick. For the tests of the two dilbit products, including the application of chemical dispersant and mineral fines (MFs), oil concentrations were highest in surface samples during the initial stages of testing as presented in Figure 5-2. These concentrations decreased over time as oil either adhered to the tank walls or was discharged through the effiuentports. Water-in-oil emulsions were visible on the seawater surface, but under these experimental conditions they were unstable since the water separated from the oil over time. The application of dispersant and MFs reduced the amount of oil detected on the seawater surface, and, in particular, chemical dispersant application alone was more effective than the other treatments.

6000

54000 5 u :t: 2000

0 0 10 20 30 40 50 60 Time (minutes)

-+-CLB -a-CLB/Disp. ...,_CLE/MF �CLB/Disp./MF

Figure 5-2. Hydrocarbon concentrations (GC-FID) in surface samples collected during the treatment of CLB under breaking waves in the wave tank.

Surface samples collected at 60 minutes were evaluated for SARA composition. Under natural conditions, the chemical composition of the dilbit products altered significantly prior to release due to weathering. The aromatics separate from the CLB with no treatment; however, the application of dispersant and MFs appears to impede this process (Figure 5-3). Similar results were found for AWB, however, not to the same degree by the application of the spill treatments. Initially, significant quantities of aromatics were detected, implying that changes in the chemical composition of the products are the result of natural processes such as evaporation and dissolution. Figure 5-4 illustrates that for AWB, the �PAH (summed parental and alkylated) concentrations are the highest within the first 30 minutes of wave tank experiments and near their lowest at a time point of 60 minutes for all oil spill treatments. For CLB, there are notably

065 higher �PAH concentrations at 60 minutes when it is treated with dispersant and/or MFs (Figure 5-5). As the low molecular weight aromatics separate from the dilbit products, the products become more dense and viscous, since they contain a greater percentage of the high molecular weight resins and asphaltenes.

During wave tank studies in non-breaking waves, oil remained on the seawater surface as a non-dispersed slick and was carried downstream where it coated parts of the tank at the seawater surface. Even under these low energy wave conditions, ''tarballs" and "tarmats" were visible on the seawater surface, but to a lesser extent compared to breaking wave conditions.

Asph

Resins

Aro

Sat

0.0% 20.0% 40.0% 60.0% 80.0% 100.0% Total Hydrocarbons

■ CLB/Disp./MF ■ CLB/Disp. ■ CLB/MF ■ CLB

Figure 5-3. Saturates, aromatics, resins and asphaltenes in extracts of surface samples collected at a time point of 60 minutes during wave tank studies of CLB.

066 30 000

25 000 �= 20 000 ::= 15 000 � � 10 000

5 000

0 000 0 5 15 30 45 60 Time (minutes)

�AWB �AWB/Disp. ..,._AWB/MF -11-AWB/Disp./MF

Figure 5-4. I;PAH concentrations in surface samples collected during various oil spill treatments of AWR

30 000

25 000 �= 20 000 ::= 15 000 � � 10 000

5 000

0 000 0 5 15 30 45 60 Time (minutes)

�CLB �CLB/Disp. _.,_CLB/MF -11-CLB/Disp./MF

Figure 5-5. I;PAH concentrations in surface samples collected during various oil spill treatments of CLR

067 5.3.4 Subsurface Water Column and Dispersant Effectiveness

5.3.4.1 In-situ Fluorescence

Without chemical dispersion, neither type of dilbit product was detected by the in-situ fluorometer. However, the addition of chemical dispersant resulted in an increase in fluorescence for both products, although fluorescence was significantlyhigher for CLB compared with A WB. The difference suggests a greater abundance of low- to mid-weight aromatic compounds in CLB than compared with A WB. The weak fluorescence in the detection of the low molecular weight aromatics associated with A WB is attributed to a lower degree of dispersion of the oil. In reference to Chapter 3, characterization of the two dilbit products revealed aromatics containing 2- to 5-ring structures. Caution should be exercised when using in-situ fluorometry to monitor unconventional oils such as diluted bitumen in the water column. The excitation and emission wavelengths of the Turner refined fuels optics are targeted towards the detection of lighter oils (i.e., shorter wavelengths). Further study needs to be conducted to determine the optimal (longer) wavelengths for the dilbit products. In addition, the influence of droplet size on the fluorescence should also be considered. Fluorometers typically respond to changes in concentrations of compounds in the dissolved phase. Since dilbit disperses into visibly larger droplets than typically observed with conventional oils, any potential bias as a result of these larger droplets on the measured fluorescenceshould be resolved.

5.3.4.2 In-situ LISST

In previous wave tank studies, conventional oil products dispersed into the water column by breaking waves in the presence of dispersant resulted in oil droplet sizes in the range of 2. 5 to 70 µm with a bimodal droplet size distribution (Li et al., 2010). These oil droplets remain dispersed and do not coalesce. However, oil droplets produced by breaking waves in the absence of dispersant are typically larger than 70 µm, have a unimodal distribution, and tend to ascend and coalesce at the surface.

For all treatments of the dilbit products in non-breaking waves, the measured particle size distributions provided no evidence of enhanced dispersion of the products. The oil simply did not disperse; no oil droplets were detected in the subsurface water column.

Figure 5-6 shows the cumulative particle size distributions located near the oil release point for the A WB over four treatments applied in breaking waves. Without dispersant, oil droplets were very large, with an average volume mean diameter (VMD) of 393 µm, and greater than 90% of total volume of all droplets were larger than 280 µm. During the experiments, large droplets could be seen in the water column. Much of the oil either resurfaced or remained on the water surface. The LISST-lO0X is only capable of detecting particles in the size range of 2.5 to 500 µm.

068 The addition of Corexit 9500 to AWB significantly altered the oil droplet size distribution, lowering the averaged VMD to 164 µm. The particle size distribution (PSD) profile also changed compared with the natural attenuation experiments, with 20% of the total volume of all droplets < 60 µm and only 15% of total particle volume > 280 µm. In addition, not only were the droplets smaller with the chemical dispersant treatment, but there were more total 1 particles in the water column with a mean total particle concentration (TPC) of 20.1 µL· L- 1 compared to a mean TPC of 4.4 µL· L- for the natural attenuation treatments.

Chemically dispersed CLB oil droplets resulted in a VMD of 71.6 µm whereas under natural attenuation, the VMD was 310.3 µm (Figure 5-6). There was also a shift in cumulative particle size distributions of CLB, with the chemically-dispersed treatments having almost 70% of the total particle volume consisting of droplets< 60 µm and less than 5% > 280 µm. Without the addition of chemical dispersants, more than 70% of the total particle volume was > 280 µm. Compared to the A WB, the lower viscosity CLB was more effectively dispersed with Corexit

9500, where there is notably a higher TPC and smaller oil droplets (Figure 5-6). These results were authenticated by higher hydrocarbon concentrations in the subsurface water column and the dynamic dispersant effectiveness data described subsequently.

Cumulative particle size distributions 12 m downstream of the oil release can be found in

Figure 5-7. For the chemically-dispersed treatment of AWB, less than 5% of the total particle volume was > 280 µm. In comparison with the CLB, very little difference was observed in the particle size distributions as the dispersed oil plume transported downstream. However, for both dilbit products, the total concentration of oil droplets in the water column decreased, likely due to dilution of the plume as it travelled horizontally. On average, the travel time of the plume over the 12 m distance was less than 10 minutes, so dissolution and coalescence would have been minimal, and hence minimal changes in particle size distributions were observed.

The effect of the addition of MF both with and without dispersant on the particle size distributions can be seen in the lower panels of Figures 5-7 for downstream LISST#2. For most of the treatments, the addition of MFs resulted in an increase in the VMD compared with the equivalent dispersant-only treatment in the top panels. A decrease in VMD when MFs were added suggests that the MFs adhered to the oil on the water surface and thus did not enter the subsurface water column. However, these data should be viewed with caution, since the LISST cannot discriminate between oil droplets, MF particles and oil-mineral aggregates in the water column.

The LISST generated unanticipated particle size distributions for CLB and AWB that are similar in shape to a unimodal distribution, with oil droplet VMDs of ~70 µm for CLB and ~160 µm for A WB, when dilbit products were treated with chemical dispersant under breaking waves (Figure 5-6). The dispersed oil droplet size patterns were dissimilar to conventional crudes. By comparison, heavy conventional oil such as IFO 180 treated with chemical dispersant under

069 similar conditions have typically produced dispersed oil droplet sizes with VMD of 60 µm with oil droplet size distribution patterns similar to unweathered medium crudes, e.g., MESA oil, a South American crude (Li et al., 2010). The droplet size distributions patterns for conventional crudesare typicallybimodal shape.

Access Western Blend - 1387 - Max TPC Point (.2 - 20 min.) w/o Dispersant w/ Dispersant

P,lean TPC = 4.4 (µUL) Mean TPC = 20.08 (µLil) Mean VMD = 393.6 (µm) Mean VMD = 164.26 (µm) a, a, 0 0

a, a, 0 0

0..,

"' 0

..0 .., 0 N N 0 0

0 0 0 0 0

LISST Size Bins. Mean Particle Size (µm) LISST Size Bins. Mean Particle Size (µm)

Cold Lake - 1387 - Max TPC Point (2 - 20 min.) w/o Dispersant w/ Dispersant

� Mean TPC • 10.23 (µLil) � Mean TPC • 73.48 (µLil) Mean VMD = 310.27 (µm) Mean VMD = 71.63 (µm) � �

a, 0 0

C: .§ 0... 0... � ,l: "' ,l:i "' 0 0 � .S! -c "' "' 0."' 0 0."' 0 2 so: 2 so: 0 � 0

:, ., ., uI 0 u 0 N � 0

0 0

0 0 0 0

LISST Size Bins. Mean Particle Size (µm) LI SST Size Bins. Mean Particle Size (µm)

Figure 5-6. Cumulative particle size distributions for AWB (top) and CLB (bottom) obtained from LISST#l (1.2 m from oil release). Each plot is an average of triplicate experiments using the particle size distribution data obtained at the time point of maximum total particle concentration within the first 20 min of the experiment (X-axis has been log transformed).

070 Cold Lake -1388 - Max TPC Point (5 -22.5 min.)

w/o Dispersant w/ Dispersant

Mean TPC = 4.1 (µUL) Mean TPC = 31.45 {µUL) Mean VMD = 264.87 (µm) Mean VMD = 79.43 (µm) .,; .,;

.,; .,;

...,; "' "' .,; .,;

N N .,; .,;

0 0 .,; .,;

LI SST Size Bins - Mean Particle Size (µm) LISST Size Bins - Mean Particle Size (µm)

w/ Kaolin w/o Dispersant w/ Kaolin & Dispersant

Mean TPC = 202.65 (µLIL) Mean TPC = 67.28 (µUL) Mean VMD = 334. 72 (µm) Mean VMD = 139.61 (µm) "' "' .,; .,;

...,; ...,;

.,;... .,;... "' "' .,; .,;

.,;

N N .,;

0 0 .,; .,;

MNmmmM�=•m�M�M�MOM -��mm��=�=�-�o MNmmmM�m•m�M�M�MoM-��=m��=�m�-�o ��MN�C��0�0� ��MN�C��C�C� NMM•��moN•�mM�Nm�MM•mQN-�OMmMm� NMM•��moN•�mM�Nm�MM•moN•�oM mMm� -----NNMM• ��m----NNNMM• -----NNMM•��=----NNNMM• LISST Size Bins - Mean Particle Size (µm) LISST Size Bins - Mean Particle Size (µm)

Figure 5-7. Cumulative particle size distributions for CLB from LISST#2 (12 m from oil release). Each plot is an average of triplicate experiments using the particle size distribution data obtained at the time point of maximum total particle concentration within the first 20 min of the experiment (X-axis has been log transformed).

5.3.4.3 Hydrocarbon Ana{yses

During the wave tank experiments, samples at vanous locations and depths were analyzed over time for hydrocarbon concentrations. In dynamic sea conditions, the mean of three replicates for three depths 12 m downstream of oil release and the effluent were plotted against time for each of the four treatments applied to the dilbit pr oducts shown in Figure 5-8. 67

071 ,,-...,4.0 CLB 0..s 82.0

:I:: 0.0 0 20 40 60 Time (minutes)

�5 cm �75 cm ....,_140 cm �Effiuent

,,-..., S 24.0 §: U 12.0 :I:: 0.0 0 20 40 60 Time (minutes)

�5 cm �75 cm ....,_140 cm �Effiuent

Time (minutes)

�5 cm �75 cm ....,_140 cm �Effiuent

,,-..., CLB/Disp./MF S 24.0 0.. U 12.0 :I:: 0.0 l«iz, i i I 0 20 40 60 Time (minutes)

�5 cm �75 cm ....,_140 cm �Effiuent

Figure 5-8. Hydrocarbon concentrations in breaking wave conditions at three depths and the effluentfor: A) CLB; B) CLB with dispersant; C) CLB with mineral fines; and D) CLB with a combination of dispersant and mineral fines.

072 The application of chemical dispersant enhanced the dispersion of both products, where hydrocarbonconcentrations were highest (~6 x) at all depths and in the oil purged from the tank. Dispersed oil droplets are influenced by current flow, and over time they are effectively diluted by clean water entering the tank. Of the two products, chemical dispersant had a greater influence on CLB, dispersing more of the product into the subsurface water column. As described previously, the in-situ LISST data also indicated that CLB was more effectively dispersed compared with A WB, supporting these findings.

MFs added as a surface active agent to enhance oil dispersion proved to be ineffective based on the limited application. A higher concentration of MFs may be more effective. Based on observations during wave tank studies, the application of MFs appeared to promote increased oil congealing, and the products appeared to have a greater tendency to stick to the sides of the tank. However, their use provides some insight on how they interact with dilbit products and aids improving the wave tank protocols, so that the natural interaction of dilbit products with sediments indigenous to Douglas Channel can be studied.

In non-breaking waves, the oil was transported as a non-dispersed surface slick downstream within minutes where it coated the tank walls and wave absorbers. Oil concentrations in the subsurface water column were consistently less than the detection limit of <1 mg/L. Application of chemical dispersant was similarly ineffective at dispersing the dilbit products. MFs were not applied under non-breaking waves.

5.3.4.4 Surface Tension

Surface tension measurements were performed on all collected seawater samples to assess the dispersion of conventional oil in the subsurface water column (King et al., 2013). In non-breaking waves, no changes were expected to be observed in seawater surface tension, since the dispersion of oil was expected to be unsatisfactory.

In breaking wave conditions, the mean surface tension measurements from three depths at sampling location "D" (14 m downstream of oil release) were plotted against time for each of the four treatments applied to the weathered dilbit products presented in Figure 5-9. Significant reductions in seawater surface tension were observed, indicating more effective dispersion of dilbit products when chemical dispersant is applied compared with natural dispersion or addition of mineral fines. Seawater surface tension decreased over time as chemically-dispersed oil droplets moved through the water column. Surface tensions eventually returned to background levels at the end of the experiment, because dispersed oil droplets were influenced by current flow and were either diluted or purged fromthe tank.

073 80.0

75.0

70.0

65.0 0 10 20 30 40 50 60 70 Time (minutes)

-a-CLB �CLB/Disp. �CLB/MF ....CLB/Disp./MF

Figure 5-9. Interfacialtension (average of all depths) for mesocosm testing of CLB under various treatment conditions.

5.3.4.5 Dynamic Dispersant Effectiveness

To evaluate dispersant effectiveness, the fraction of oil entrained in the water column of the wave tank had to be determined. A large number of discrete samples were collected, as indicated previously in section 5.2.1, to directly measure the oil dispersed in the water column. With the wave tank in flow-through operation, assessments of dispersant effectiveness also included the concentration of oil in the effiuent purged from the tank. Dynamic dispersant effectiveness (DDE) can be used to evaluate the fate and transport of spilled dilbit, including the application of oil spill treating agents under natural sea states and environmental conditions (Li et al., 2010). Since dispersion of dilbit products was ineffective in non-breaking waves, DDE was applied only to treatments of spilled dilbit products in breaking waves. DDE can therefore provide information on the best response options encountered in actual spills based on conditions simulated in the wave tank.

DDE (equation 1) over the duration of an entire experiment can be evaluated by computing the fraction of dispersed oil purged from the wave tank (%Foil) and the residual dispersed oil in the water column (%Dwc) at the end of each experiment:

%DDE =%Pou+ %Dwc (1)

Further details on DDE calculation can be found in Li et al., 2010. Oil was applied quantitatively; therefore, the distribution of the entire amount of oil released can be estimated. The assumption here is that total oil released (T0il; 100%) is the summation of %DDE and the percentage of oil on the water surface or stuck to the walls of the tank (%Soi/). Theonly unknown 70

074 is %Soil (equation 2), which can be calculated as follows:

%Sazt = �u - %DDE (2)

The averaged hydrocarbon values for all depths at each sampling location (A, B, C and

D) were used to generate %DDE values. The average values of triplicate trials for %Sail, %Pail and %Dwc for all treatments were plotted for each dilbit product (Figure 5-10). Analysis of Variance (ANOVA evaluating oil dispersion; Microsoft Excel 2010) illustrates the effects of dispersant, MFs and the combination of MFs and dispersant on the fate and transport of CLB and AWB in breaking wave conditions. A p-value ::=; 0.1 provides a 90% confidence that the mean values are statistically different. Evaluation of oil spill treatment agents illustrated that both dispersant and MPs/dispersant influenced CLB dispersion (p-values ::=;0.1 for 90% confidence interval), where dispersant only had an effect on AWB (p-value ::=;0.1 for a 90% confidence interval) dispersion compared with the other spill treating options under similar conditions (Tables 5-2 and 5-3). At seawater temperatures of 8.3 ± l.3°C, the estimated %DDE for Corexit 9500 applied to CLB and AWB was 45% and 30%, respectively. However, a greater percentage of released oil remained as a non-dispersed oil slick on the seawater surface. The data generated represents the testing of two dilbit products under a limited range of hydrodynamic and environmental conditions.

CLB/disp/MF

CLB/MF

CLB/disp.

CLB

A WB/disp/MF

AWB/MF

AWB/disp.

AWB

0.0% 20.0% 40.0% 60.0% 80.0% 100.0%

Figure 5-10. Dynamic dispersion effectiveness (n=3) for AWB and CLB in breaking wave conditions using = = = various treatment options (P0ll oil purged from the wave tank, D..., dispersed oil in the water column, S0ll oil remaining on the water surfaceor adhered to walls of the tank).

075 Table 5-2. Analysis of variance comparing the effects of dispersant, mineral fines and mineral fines/dispersant to the natural dispersion of CLB under breaking waves.

Treatment n ss df MS F v-value F crit Criteria Dispersant 3 2541 2541 9.0 0.04 4.5 Accept; p-.5:.0.1 MF 3 18 18 1.3 0.31 4.5 Reject; p>0.1 MF/Dispersant 3 1324 1324 4.5 0.10 4.5 Accept; p<0 1

Table 5-3. Analysis of variance comparing the effects of dispersant, mineral fines and mineral fines/dispersantto the natural dispersion of AWB under breaking waves.

Treatment n ss df MS F p-value F crit Criteria Dispersant 3 942 942 4.5 0.10 4.5 p-.S.0.1 MF 3 6.2 6.2 0.11 0.75 4.5 Reject; p>0.1 MF/Dispersant 3 77 77 0.72 0.44 4.5 Reject; p>0.1

076 6.0 Conclusions

A review of the scientific literature revealed knowledge gaps on the physical and chemical properties of dilbit and synbit products; their fate, behaviour and transport in the marine environment; and the application of oil spill response technologies to enhance the natural attenuation of the products.

Two diluted bitumen products, which are representative of high-volume commercial pipeline traffic in Canada, A WB and CLB, were evaluated using a variety of chemical and physical tests. The physical and chemical properties for unaltered and synthetically weathered dilbit products are presented. As seen in Chapter 4, the measured properties of the synthetically weathered products indicate that evaporation alone is likely not sufficient to cause sinking of either product in marine environments, even at temperatures approaching freezing. Mixing of the products with seawater led to water-uptake that increased the density of the resulting mixture; however, the mixtures remained buoyant in all cases.

The laboratory studies described in Chapter 3 and 4 provide essential information on dilbit physical and chemical properties, product weathering, and fate and behaviour under various controlled conditions on a small scale. These small-scale lab studies provide a quick, initial assessment of factors that affect oil behaviour and fate. The information forms the foundation to further investigate the behaviour and fate of oil products in sea states and environmental conditions that can be simulated in a wave tank facility.

As discussed in Chapter 4, the buoyancy behaviour of the A WB and CLB products in marine conditions depends most strongly on the presence of medium-to-fine sediment in the water column. Evaporative weathering alone and evaporative and photo-oxidative weathering in combination all resulted in products that were buoyant in marine conditions. Mixing with water generally increased the density of the products, but all oils tested remained buoyant in seawater even when saturated with water. When mixed with fine- and moderate-sized sediments, however, the fresh to moderately weathered diluted bitumen products sank in saltwater. Interestingly, when mixed with the same fine sediments at high degrees of evaporative weathering, the more viscous oil did not mix as well with either fineor medium sediment, and discrete floating tarballs began to form. The work illustrates the first observation of the initial stages of tarball formation for highly evaporated dilbit. When mixed with coarse sediment, the lower-viscosity fresh and moderately-weathered fractions of both dilbit products did not readily take up silica sand, and instead formed mostly floating entrained-water-in-oil states, similarto those formedby seawater oil mixtures with no sediment present. However, some large sand-encrusted blobs of oil were observed to sink in the sand-mixing tests.

This work demonstrates that, in waters where fine- to moderate-sized sediment is present, these oils are at risk to sink, when there is a high degree of mixing energy available. However, 73

077 the effects of different mixing regimes, including current flow, on oil-sediment interactions have not been examined in the present work. Comparisons to meso-scale testing in lower mixing energies by other researchers have revealed some differences between, for example, water uptake by oils. Testing in the wave tank described in Chapter 5, moderate mixing of the oil sediment aggregates, resulted in a suspension of the materials. Available mixing energy factors seem to have an influence on the fate of the formed oil-sediment aggregates. While the present work illustrates some of the forms that these oils may possibly adopt following a spill, more work is needed to understand the mechanisms and rates of formation of these states, and to understand the factors that govern the transitions between these fates.

In Chapter 5, studies were conducted in a flow-through wave tank facility to examine the fate, behaviour and transport of oil at the meso-scale, as well as the effectiveness of selected oil spill treating agents in breaking and non-breaking waves with current flowin seawater at average temperatures of 8.3 ± l.3°C. In non-breaking waves, the dilbit products were not influenced by oil spill treating agents, individually or in combination, and remained as a non-dispersed slick on the seawater surface.

In breaking wave conditions, untreated dilbit entered the subsurface water column as large droplets that quickly resurfaced and coalesced into a surface oil slick. Natural processes acting on the slick significantly altered the chemical composition of the products, forming visibly weathered fractions including unstable water-in-oil emulsions, tarballs and tarmats on the seawater surface. Application of chemical dispersant caused partial dispersion of the products in breaking waves at an average seawater temperature of 8.3 ± l.3°C. Chemical dispersant application was more effective in dispersing the lower-viscosity CLB. These findings were supported by particle size distributions; however, the observed oil droplet size distributions were unimodal, distinctly different from the bimodal oil droplet distributions for dispersion of conventional oils. Even though the dilbit products were partially dispersed in breaking waves, a greater percentage of the released oil persisted as a non-dispersed oil slick at the seawater surface.

The in-situ fluorescence instrument was challenged when tracking the chemically dispersed AWB, but showed some promise in monitoring chemically-dispersed CLB in the water column. The fluorometer is a field unit that records changes in fluorescence compared with background levels (prior to oil release). The instrument used in this study has a fixed wavelength, and it is limited to detecting aromatics with 1- to 3-ring structures. As outlined in Chapter 3, dilbit products contain aromatics with 1- to 5-ring structures. Deploying fluorometers with different fixed wavelengths to cover the full spectrum of aromatics will provide more information on the fateof dilbit products in marine settings.

The question of whether dilbit products spilled in the marine environment will float or sink depends on their exposure to a number of natural processes and the duration of exposure.

078 These wave tank studies suggest that if these two dilbit products lose ~6 to 9% of their mass by natural processes such as evaporation, they will remain floating, and partial movement into the water column with the application of dispersant may be possible under the limited environmental conditions outlined in this study.

In previous testing (Li et al., 2009a; Li et al., 2009b; Li et al., 2010), it was shown that a variety of oil types are effectively dispersed in a dynamic environment within the first hour of their release and that temperature can have an effect on oil spill treating agent effectiveness. Longer test times in the current study with dilbit products could not be considered, as they adhered to the walls and wave absorbers of the tank restricting further evaluation. These were unexpected observations that did not transpire during previous work with conventional oils. Effective dispersion of oil products produces small oil droplets <70 µm in size. These small droplets are buoyant in the water column where they are diluted by current and spatially distributed. In addition, these small droplets provide an ideal surface area for microbes to attach themselves, and the carbon associated with the oil droplets provides a source of energy for them. This process of dispersing oil potentially encourages other natural processes such as biodegradation of the spilled oil. The chemical dispersion of oil may be applicable to achieve a "Net Environment Benefit" where all available spill response options need to be considered. The oil spill treating agent ( chemical dispersant) and mineral fines used in this study were either ineffective or partially effective when applied to the dilbit products under these limited conditions. The physical properties (e.g., density, viscosity and adhesiveness) of these products limit the effectiveness of currently-available spill treating agents, thus restricting remediation and potentially contributing to the persistence of the products in marine environments where seawater temperature is <8°C.

The wave tank studies outlined in this report were conducted under limited conditions; partiallycovering the temperature ranges listed in Chapters 3 and 4 and weathered dilbit products similar to "Wl" .. Further work is required to determine the influence of other temperature ranges, in particular in warmer waters, on these products and how they influence the effectiveness of spill treating agents. Since the treating agents applied to the dilbit products under these conditions were ineffective to partially effective, there is no need to test products weathered beyond "Wl." Treatment of weathered dilbit products beyond "Wl" would require the development of new spill treating agents.

079 7.0 Research Initiatives

7.1 Identified Knowledge Gaps from the Literature Review

From a review of the literature, a number of knowledge gaps related to dilbit and synbit were identified, including:

1. The rate of diluent evaporation could be difficult to predict during a spill. The applicability of oil spill evaporation models has not been assessed for these products. 2. How weathering of dilbit or synbit affects its tendency to float, sink or become suspended in the water column is not well known. 3. How long it might take for the oil to leave the surface in a dilbit or synbit spill is unknown, yet this is significantfor the window of opportunity during spill response. 4. Research regarding how bitumen products will further biodegrade in the environment is insufficient. 5. A variety of bitumen products should be tested, including application of existing oil spill treating agents in natural sea states and environmental conditions, since diluted bitumen properties can be diverse.

7.2 Gaps Addressed Under This Research Initiative

Based on the gaps identified in the literature review (section 7.1 ), the following items were addressed in the present work:

7.1, Bullets 1, 2 and 3:

The physical and chemical properties were characterized for two winter dilbit blends. The oil products showed some similarities, but there were also differences in physical characteristics and chemical composition, in particular in the amount of condensate blended with the bitumen and between the bitumen feedstocks themselves. Preliminary results discussed in Chapter 3 suggest that some models of evaporation may not be very suitable to diluted bitumen products. Differences in the distillation data and in composition data for the saturate, aromatic, resin and asphaltene groups were apparent for the two dilbit types. The oil products were weathered to various degrees and their physical properties were assessed. The dilbit products and their various weathered samples were evaluated under limited environmental conditions. In their most weathered forms, the products were found to float on seawater; however, interactions with sediments can change the behaviour and fate of the products where partial sinking was observed and some suspended tarball formation occurred.

7.1, Bullet 5:

Wave tank studies were conducted on two slightly weathered dilbit products (similar to "W l" in Chapter 3) spilled on seawater, including the application of an oil spill treating agent 76

080 and sediment. Thetwo dilbit products were tested in breaking and non-breaking wave conditions and in seawater at an average temperature of 8.3 ±1.3 °C. The products applied to seawater float, and the application of chemical dispersant showed partial dispersion of the oil products under these limited test conditions. Differences in the products' behaviour and fate included greater effectiveness for chemically-enhanced dispersion of CLB compared to AWB. Overall, under these limited conditions, the products were observed to be very sticky, and they adhered to the tank walls and absorbers, thus limiting the dispersion of the oil products.

Finally, the findings of this work should be integrated into fate and behaviour models and, eventually, incorporated into large-scale circulation models.

7 .3 Gaps Identified from These Studies and Future Research Initiatives

The physical and chemical properties of the two dilbit products examined show some dissimilarity. Laboratory testing of the products and their weathered samples showed differences in their behaviour and fate when evaporated (rates of evaporation), mixed with saltwater (formations of emulsions versus entrained water mixtures), exposed to ionizing light (water uptake when forming oil-saltwater mixtures) and in how they interacted with sediments. There is lack of information on the nature of sediment interaction with dilbit products, the rate of tarball formationand mixing energy to sediment loadings required to form tarballs. In Chapter 4, it was shown that when mixed with fine- and moderate-sized sediments, the fresh-to-moderately weathered dilbit products dispersed and sank. These studies were conducted in high-energy mixing flasks in a laboratory, so it is uncertain whether dispersed products would be suspended or sink under particular natural conditions. Future work should investigate the interaction of dilbit products with suspended sediments on a larger scale and under different mixing energy regimes to determine the fate and transport of oil droplets influenced by sediment interactions (oil-sediment aggregate formation) in natural sea states and environmental conditions. Further, a range of sediment types, including natural sediments, and sediments incorporating organic material should also be included in anyfuture testing.

Since access to products weathered in the natural environment is difficult, oil products used in these studies were synthetically weathered. It is not well understood how synthetically weathered oil products relate to various degrees of weathering of dilbit products in the natural environment based on hours, days and weeks. This point to the need for an enhanced test facility in which oil products can be exposed to controlled weathering conditions over an extended period of time, including control of temperature, illumination, and wind and wave-mixing conditions.

Laboratory studies have resulted in the measurement of physical and chemical properties of two representative dilbit products under different temperature conditions. The large-scale wave tankstudies were limited to 8.3 ± l.3°C. Further meso-scale wave tank studies are required for other temperature ranges, to cover the whole range of surface sea temperatures experienced 77

081 in the Canadian marine environment. These studies are important in order to determine how warmer and colder seawater temperatures will affect the fate and behaviour of dilbit products under natural sea states. In addition, there is a need for new spill treating agents that are effective on dilbit products in seawater at temperatures below 8°C. Under these preliminary test conditions, the interaction of dilbit products with microbes and their potential to biodegrade these products have not been considered. This gap was also identifiedin the literature review.

The Government of Canada is working closely with academia in an effort to assess the biological effects of a dilbit spill as well as studying the effectiveness of available spill treating agents. Biological effects studies will evolve as research continues toward assessing the behaviour and fateof dilbit spilled at sea.

In order to adequately characterize the behaviour and fate of the two winter dilbit products, further studies are required:

7.3.1 Laboratory Dilbit Fate and Behaviour Studies

• The preliminary buoyancy and behaviour studies on dilbit products summarized in the present work will be applied to the design of a future set of experiments to examine the nature of these interactions, to understand the rates of formation, the effects of sunlight and ionizing light exposure, the thresholds of mixing energy/sea state and sediment types and concentrations necessary to pose risks of sinking or tarball formation. • Work will also be undertaken to develop new methods for understanding oil weathering behaviour, with in-situ control of evaporation, UV exposure, shear-mixing energies and temperatures. It is anticipated that such facilities and tools will allow for improved simulation of real-world spill fates and behaviour, and also allow for long-term in-situ weathering experiments.

7.3.2 Meso-Scale Studies and In-situ Field Work

• Evaluate seawater temperatures above 15 °C on spill treating agent effectiveness and their influence on the behaviour andfate of the winter dilbit products (wave tank studies); • Evaluate existing laboratory assessment methods for spill-treating effectiveness screening (e.g., the swirling flask test and the baffled flask test), and compare with the results of the wave tank studies; • Natural weathering of the winter products to establish a link with degrees of weathering in relationship to the number of days at sea and to evaluate how long the winter dilbit products will remain on the seawater surface (meso-scale and field studies); • In-situ field experiments to assess the biodegradation rates of the winter dilbit products and the products treated with dispersant to address Bullet 4 under gaps identified in the literature review and this study, both in seawater and on marine shorelines;

082 • Over the next two years, the initial development of algorithms to predict the fate and transport of diluted bitumen products in the marine environment; and • Interaction of these winter dilbit products with suspended sediments in natural sea states and environmental conditions at the meso-scale in a wave tank to further evaluate the findings from the lab studies in Chapter 4 of this report.

083 8.0 References

Aijiolaiya, L.O., Hill, P.S., Khelifa, A., Islam, R.M., and Lee, K., 2006. Laboratory Investigation of the Effects of Mineral Size and Concentration on the Formation of Oil-Mineral Aggregates, Marine Pollution Bulletin, 52, pp. 920-927. Alberta Energy, 2011. Website: www.energy.alberta.ca/OilSands/791.asp, accessed August 2013. Alberta Geological Survey, 2013. Website: www.ags.gov.ab.ca/energy/oilsands/index.html. accessed September 2013. Amos, C., 1996. Siliciclastic tidal flats, in: Perillo, G.M.E. (Ed.), Geomorphology and sedimentology of estuaries, 2nd Edition. Elsevier, Amsterdam, pp. 273-306. ASL Environmental Sciences Inc., 2010. Marine Physical Environmental in the CCAA Technical Data Report. Prepared for: NorthernGateway Pipelines Inc., Calgary, AB. ASTM D 97, Standard test method for pour point of petroleum products, ASTM International, West Conshohocken, PA, 2012. ASTM D2007, Standard Test Method for Characteristic Groups in Rubber Extender and Processing Oils and Other Petroleum-Derived Oils by the Clay-Gel Absorption Chromatographic Method, ASTM International, West Conshohocken, PA, 2011. ASTM D 4294, Standard Test Method for Sulfur in Petroleum and Petroleum Products by Energy Dispersive X-ray Fluorescence Spectrometry, ASTM International, West Conshohocken, PA, 2010. ASTM D 5002, Standard test method for density and relative density of crude oils by digital density analyzer, ASTM International, West Conshohocken, PA, 2010. ASTM D7169, Standard Test Method for Boiling Point Distribution of Samples with Residues Such as Crude Oils and Atmospheric and Vacuum Residues by High Temperature Gas Chromatography, ASTM International, West Conshohocken, PA, 2011. ASTM G 173-03, Standard Tables for Reference Solar Spectral Irradiances: Direct Normal and Hemispherical on 37° Tilted Smface, ASTM International, West Conshohocken, PA, 2003. Bernabeu, AM., Fernandez-Fernandez, S., Bouchette, F., Rey, D., Arcos, A., Bayona, J.M., and Albaiges, J., 2013. Recurrent arrival of oil to Galician coast: The final step of the Prestige deep oil spill, Journalof Hazardous Materials, 250-251, pp. 82-90. Belore, R., 2010. Technical Data Report: Properties and Fate of Hydrocarbons Associated with Hypothetical Spills at the Marine Terminal and in the Confined Channel Assessment Area, SLRoss Environmental Research Ltd., Ottawa, Ontario, 40 p. ENG exhibit number Bl6-31. Brown, J., Fuentes, H., Jaffe, R., and Tsihrintzis, V., 1995. Comparative evaluation of physical and chemical fate processes of Orimulsion and fuel oil No. 6 in the Tampa Bay marine environment. In: Comparative Ecological Risk Assessment-Technical Support Document for the Comparison of the Ecological Risks to the Tampa Bay Ecosystem from Spills of 80

084 Fuel Oil #6 and Orimulsion-Final Report, Ault, J. S., M. A. Harwell, and V. Myers, (eds.). Miami, Florida: University of Miami, Center for Marine and Environmental Analyses, Rosenstiel School of Marine and Atmospheric Science. Bobra, M., and Tennyson, E.J., 1989. Photooxidation of Petroleum, in the Proceedings of the Twelfth Arctic Marine Oilspill Program Technical Seminar, Environment Canada, Ottawa, Ontario, pp. 129-147. Chapman, H., Purnell, K., Law, R.J., and Kirby, M.F., 2007. The use of chemical dispersants to combat oil spills at sea: A review of practice and research needs in Europe, Marine Pollution Bulletin, 54(7), pp. 827-838. Cheng, P., Li, D., Boruvka, L., Rotenberg, Y., and Neumann, A.W., 1990. Automation of axisymmetric drop shape analysis for measurements of interfacial tensions and contact angles, Colloids and Surfaces, 43(2), pp. 151-167. Clark, B., Parsons, J., Yen, C., Ahier, B., Alexander, J., and Mackay, D., 1987. A Study of Factors Influencing Oil Submergence, EE-90, Environment Canada, Ottawa, Ontario, 77 p. Clark, J., Becker, K., Venosa, A., and Lewis, A., 2005. Assessing dispersant effectiveness for heavy fuel oils using small-scale laboratory tests, in Proceedings of the International Oil Spill Conference, American Petroleum Institute, Washington, D.C., pp. 59-63. Colcomb, K., Salt, D., Peddar, M., and Lewis, A., 2005. Determination of the limiting oil viscosity for chemical dispersion at sea, in Proceedings of the International Oil Spill Conference, American Petroleum Institute, Washington, D.C., pp. 53-58. Cole, M., King, T., and Lee, K., 2007. Analytical technique for extracting hydrocarbons from water using sample container as extraction vessel in combination with a roller apparatus. Canadian Technical Report of Fisheries and Aquatic Sciences 2733, pp. 1-12. Crosby, S., Fay, R., Groark, C., Kani, A., Smith, J. R., and Sullivan, T., 2013. Transporting Alberta's Oil Sands Products: Defining the Issues and Assessing the Risks. 180 p. Website: www.crrc.unh.edu/workshops/oil sands washington/Final%20NOAA %20Oil%20Sands %20Report 03 22.pdf. Crude Quality Inc., 2013. Website: crudemonitor.ca, www.crudemonitor.ca/home.php, accessed September 2013. Diez, S., Jover, E., Bayona, J.M., Albaiges, J., 2007. Prestige oil spill. III. Fate of a heavy oil in the marine environment, Environmental Science and Technology, 41(9), pp. 3075-3082. DFO, Department ofFisheries and Oceans, Coastal Shallow Water Temperature Climatology for Atlantic Canada, www2.mar.dfo- mpo.gc.ca/science/ocean/coastal temperature/coastal temperature.html, accessed November 2013. DFO, Department of Fisheries and Oceans, Sea Surface Temperature in the Gulf of Alaska, www.pac.dfo-mpo.gc.ca/science/oceans/data-donnees/sst-tsm/index-eng.html, accessed November 2013.

085 Drelich, J., Fang, Ch., and White, C.L., 2002. Measurement of interfacial tension in fluid-fluid systems. Encyclopedia of Colloid Science, pp. 3152-3165. Drennan, W.M., Donelan, M.A., Terray, E.A., and Katsaros, KB. 1996. Oceanic turbulence dissipation measurements in SWADE. Journal of Physical Oceanography, 26(5), pp. 808- 815. ENG, 2011. Enbridge Northern Gateway, 2011, Information Requests Round # 1 to the Applicant Government of Canada, ENG JRP Exhibit #B41-4, 246 p. Environment Canada, 2006. Oil Properties Database. www.etc- cte.ec.gc.ca/databases/OilProperties/oil prop e.html accessed August 2013. EPA, 2013. Dredging Begins on Kalamazoo River, www.epa.gov/enbridgespill/pdfs/enbridge fs 201308.pdf, accessed October 2013. ERCB, 2013. Energy Resources Conservation Board, 2013. Alberta's Energy Reserves 2012 and Supply/DemandsOutlook 2013-2022, ST98-2013, p3-l. www.aer.ca/documents/sts/ST98/ST98-2013.pdf accessed November 2013. Fieldhouse, B.G., Hollebone, B.P., Singh, N.R., Tong, T.S., and Mullin, J.V., 2010. Artificial Weathering of Oils by Rotary Evaporator, Proceedings of the Thirty-third AMOP Technical Seminar on Environmental Contamination and Response, Environment Canada, Ottawa,Ontario, pp. 159-180. Fingas, M. 1999. Evaporation of oil spills: development and implementation of new prediction methodology, in Proceedings of the International Oil Spill Conference, American Petroleum Institute, Washington, D.C., pp. 281-287. Fingas, M., 2004. Modeling evaporation using models that are not boundary-layer regulated Journal of Hazardous Materials, 107(1-2), pp. 27-36. Fingas, M., and Fieldhouse, B, 2006. A Review of Knowledge on Water-in-oil Emulsions, in Proceedings of the Twenty-ninth Arctic and Marine Oilspill Program Technical Seminar, Environment Canada,Ottawa, Ontario, pp. 1-56. Fingas, M., Fieldhouse, B., Wang, Z., and Landriault, M., 2005. The dynamics of Orimulsion in water with varying energy, salinity and temperature, in Proceedings of the International Oil Spill Conference, American Petroleum Institute, Washington, D.C., pp. 11408-11411. Fingas, M., Hollebone, B., Wang, Z., Smith, P., and MacKay, R., 2003. The properties of heavy oils and Orimulsion: Another look, in Proceedings of the Twenty-sixth Arctic and Marine Oilspill Program Technical Seminar, Environment Canada, Ottawa, Ontario, pp. 43-62. Galbraith, P.S., Larouche, P., Chasse, J., Petrie, B., 2012. Sea-surface temperature in relation to air temperature in the Gulf of St. Lawrence: Interdecadal variability andlong term trends, Deep-Sea Research II: Topical Studies in Oceanography,77-80 , pp.10-20 Guitart, C., Prickers, P., Horrilla-Carballo, J., Law, R.J., and Readman, J.W. Characterization of sea surface chemical contamination after shipping accidents. Environmental Science and Technology, 42(7), pp. 2275-2282. Harper, J.R. and Kory, M., 1997. "Orimulsion: Sediment Interaction Experiments,", Environment Canada, Ottawa, ON, 45 p.

086 Harper, J.R. and Ward, S., 2003. "Adhesion and Removal of Orimulsion Bitumen from Rock Surfaces," Environment Canada, Edmonton, Alberta, 51 p. Hollebone, B., Wang, Z., Landriault, M., and Smith, P., 2003. A New Method for the Determination of the Hydrocarbon Groups in Oils: Saturates, Aromatics Resins and Asphaltenes (SARA), in Proceedings of the Twenty-sixth Arctic and Marine Oilspill Program Technical Seminar, Environment Canada, Ottawa, Ontario, pp. 31-42. Hollebone, B., 2010. "Measurement of Oil Properties," in Oil Spill Science and Technology, Ed. M. Fingas, Elsevier Inc., Oxford, UK, pp. 63-85. King, T., Clyburne, J.A.C., Lee, K. and Robinson, B., 2013. Interfacial film formation: Influence on oil spreading rates in lab basin tests and dispersant effectiveness testing in a wave tank. Marine Pollution Bulletin, 73(1-2), pp. 83-91. Kostaschuk, R.A., Stephan, B.A., and Luternauer, J.L., 1993. Suspended sediment concentration in a buoyant plume: Fraser River, Canada, Geo-Marine Letters, 13(3), pp. 165-171. Jokuty P, Whiticar S, Wang Z, Doe K, Fieldhouse B, and Fingas M. 1999. Orimulsion-400: A comparative study. Report EE-160. Environment Canada, Ottawa, ON. Lee, K., Bugden, J., Cobanli, S., King, T., McIntyre, C., Robinson, B., Ryan, S., Wohlgeschaffen, G., 2012. UV- Epifluorescence Microscopy Analysis of Sediments Recovered from the Kalamazoo River, U.S. EPA Kalamazoo Administrative Record, document #1277, www.epa.gov/enbridgespill/ar/enbridge-AR-1277.pdf, accessed November 2013. Li, Z., Lee, K., King, T., Boufadel, M. C., and Venosa, A. D., 2008. Assessment of chemical dispersant effectiveness in a wave tank under regular non-breaking and breaking wave conditions. Marine Pollution Bulletin, 56(5), pp. 903-912. Li, Z., Lee, K., King, T., Boufadel, M. C., and Venosa, A. D., 2009a. Evaluating crude oil chemical dispersion efficacy in a flow-through wave tank under regular non- breaking wave and breaking wave conditions. Marine Pollution Bulletin, 58(5), pp. 735-744. Li, Z., Lee, K., King, T., Kepkay, P., Boufadel, M., and Venosa, A., 2009b. Evaluating chemically dispersant efficacy in a wave tank: dispersant effectiveness as a function of energy dissipation rate. Environmental Engineering Science, 26(6), pp. 1139-1148. Li, Z., Lee, K., King, T., Boufadel, M. C., and Venosa, A. D., 2010. Effects of temperature and wave conditions on chemical dispersion efficacy of heavy fuel oil in an experimental flow-through wave tank. Marine Pollution Bulletin, 60(9), pp. 1550-1559. Lunel, T., and Davies, L., 2001. Response to Bunker Fuel Oil: The Options, in Proceedings of the International Oil Spill Conference, American Petroleum Institute, Washington, D.C., pp. 597-603. Maki, H., and Sasaki, T., 1997. Analytical method of crude oil and characterization of spilled heavy oil. Journal JapanSociety on Water Environment, 20(9), pp. 639-642. Michel, J., 2010. "Submerged Oil" in in Oil Spill Science and Technology, Ed. M. Fingas, Elsevier Inc., Oxford, UK, pp. 959-981.

087 NEB, 2012. National Energy Board, Enbridge Northern Gateway Project Joint Review Panel, http://gatewaypanel.review-examen.gc.ca/clf-nsi/prtcptngprcss/rlsttmntmp-eng.html, accessed November 2013. NRC, 1999, Spills of Nonfloating Oils: Risk and Response, National Academy of Science, National Research Council, Washington, D.C., 88 p. Ostazeski, S. A., Macomber, S. C., Roberts, L. G., Uhler, A. D., Bitting, K. R., and Hiltabrand, R., 1997. The Environmental Behavior of Orimulsion Spilled on Water, in Proceedings of the International Oil Spill Conference, American Petroleum Institute, Washington, D.C., Vol. 1997, pp. 469-477. Owens, E.H., Taylor, E., Humphrey, B., 2008. The persistence and character of stranded oil on coarse-sediment beaches, Marine Pollution Bulletin, 56(1), pp. 14-26. Owens, E.H. 2010. Shoreline response and long-term oil behaviour studies following the 1970 "Arrow" spill in Chedabucto Bay, NS, in Proceedings of the Thirty-third AMOP Technical Seminar on Environmental Contamination and Response, Environment Canada, Ottawa, Ontario, pp. 207-221. Read, J., and Whiteoak, D., 2003. The Shell Bitumen Handbook, 5th edition, Thomas Telford Publishing, London, UK, 460 p. SL Ross Environmental Research Ltd., 2013. Meso-scale Weathering of Cold Lake Bitumen/Condensate Blend, Ottawa, Ontario, 2012, 27 p., ENG JRP Exhibit #Bl93-2. Spieweck, F., and Bettin, H., 1992. Review: Solid and liquid density determination, Technisches Messen, 59, pp. 285-292. Stevens, L., and Roberts, J., 2003. Dispersant effectiveness on heavy fuel oil and crude oil in New Zealand, in Proceedings of the International Oil Spill Conference, American Petroleum Institute, Washington, D.C., Vol. 2003, No. 1, pp. 509-513. Stantec Consulting Ltd., 2012a. Technical Data Report: Ecological and Human Health Risk Assessment for Pipeline Spills, Calgary, Alberta, ENG JRP Exhibit #B80-2, 94 p. Stantec Consulting Ltd., 2012b. Summary of Clean up and Effects of the 2007 Spill of Oil from the Trans Mountain Pipeline to Burrard Inlet, www.transmountain.com/uploads/pages /13 7 4960812-2012-Summary-2007 -Spill Clean-U p---Effects-REV2.pdf accessed October 2013. Stout, S. A. 1999. Predicting the Behavior of Orimulsion Spilled on Water. U.S. Department of Transportation, Project No. 4120.ll/UDI283, U.S. Coast Guard, Groton, Connecticut. Stronach, J., Wang, E., Draho, B., and Miguez, T., 2010. Wind observations in Douglas Channel, Squally Channel, and Caamano. Hayco Technical Data Report, www.ceaa.gc.ca/050/documents staticpost/cearref 21799/2431/Wind lof3.pdf accessed November, 2013. Terray, E.A., Donelan, M.A., Agrawal, Y.C., Drennan, W.M., Kahman, K.K., Williams, A.J., Hwang, P.A., and Kitaigorodskii, S.A., 1996. Estimate of kinetic energy dissipation under breaking waves. Journal of Physical Oceanography,26(5), pp. 792-807.

088 Transport Canada, 1995. Response Organizations Standards, Marine Safety Publication TP 12401. www.tc.gc.ca/ eng/marinesafety/tp-tp1240 l-menu-2162.htm accessed October 2013. Transport Canada, 2013. Website: www. tc. gc. ca/ eng/mediaroom/releases-20 l 3-h03 le-7089. htm accessed August 2013. TSB, 2007. Transportation Safety Board of Canada Pipeline Investigation Report P07H0040. www.ts b. gc.ca/en g/rapports-reports/pipeline/2007/ p07h0040/p07h0040.asp accessed August 2013. Trudel, B., Belore, R., Guarino, A., Lewis, A., and Mullin, J., 2005. Determining the viscosity limits for effective chemical dispersion: relating Ohmsett results to those from tests at sea, in Proceedings of the International Oil Spill Conference, American Petroleum Institute, Washington, D.C., pp. 71-76. Uncles, R.J., Stephens, and J.A, Smith, R.E., 2002. The dependence of estuarine turbidity on tidal intrusion length, tidal range and residence time, Continental Shelf Research, 22, pp. 1835-1856. UNESCO, 1981. Tenth report of the Joint Panel on Oceanographic Tables and standards, UNESCO Technical Papers in Marine Science 36, United Nations Educational, Scientific and Cultural Organization, Division of Marine Sciences, Paris, France, 28 p. US Environmental Protection Services, 1996. EPA Method 8270C: Semivolatile Organic Compounds by Gas Chromatography/Mass Spectrometry. Revision 3, pp. 1-54. US Geological Survey 2006. National Assessment of Oil and Gas Fact Sheet: Natural Bitumen Resources of the United States (Vol. 2006�3133). Wang, Z. D., Fingas, M. and Li, K., 1994. Fractionation of ASMB Oil, Identification and Quantitation of Aliphatic Aromatic and Biomarker Compounds by GC/FID and GC/MSD (Parts I and II), Journalof ChromatographicScience, Vol. 32, pp. 361-382. Wang Z, and Fingas, M. 1996. Separation and characterization of petroleum hydrocarbons and surfactant in Orimulsion dispersion samples. Environ. Sci. Technol., Vol. 30, pp. 3351- 3361. WEC, 2010. World Energy Council "2010 Survey of Energy Resources. Natural Bitumen and Extra-Heavy Oil", Chapter 4, pp. 123-150. energy.usgs .gov/portals/0/Rooms/economics/text/WEC1 0NB EHO.pdf, accessed November 2013. Wickley-Olsen, E., Boufadel, M., King, T., Li, Z., Lee, K., and Venosa, A., 2008. Regular and breaking waves in a wave tank for dispersion effectiveness testing, in the Proceedings of the International Oil Spill Conference, American Petroleum Institute, Washington, D.C., pp. 499-508. Yang, C., Wang, Z., Yang, Z., Hollebone, B.P., Brown, C.E., Landriault, M. and Fieldhouse, B., 2011. Chemical Fingerprints of Alberta Oil Sands and Related Petroleum Products, Environmental Forensics, 12(2), pp. 173-188.

089 e-ISSN 2236-1057 - doi:10.5597/lajam00070 http://dx.doi.org/10.5597/lajam00070 LAJAM 4(1): 55-60, January/June 2005 ISSN 1676-7497

WHALE MORTALITY FROM SHIP STRIKES IN ECUADOR AND WEST AFRICA

FERNANDO FÉLIX1 AND KOEN VAN WAEREBEEK2

ABSTRACT: We document two unusual cases of a Bryde’s and a sei whale struck by container cargo vessels and draped over the bow bulb, respectively in the Southeast Pacific and the Eastern Tropical Atlantic. The 207m-length P&O Nedlloyd Pantanal collided with an adult Bryde’s whale Balaenoptera edeni in the southern Gulf of Guayaquil, Ecuador, on 10 December 2004. The whale showed massive dermal hematoma indicating that it was alive when struck. Similarly, the container ship OSNA Bruck arrived at Dakar port, Senegal, on 19 March 1998 with the fresh carcass of a juvenile sei whale B. borealis on its bow bulb. The collision occurred between Las Palmas, Gran Canaria and Dakar, Senegal. Freshly dead balaenopterids can hardly be picked up by ships since they sink at death and do not float until decomposition and bloating sets in. A review of previous cases show that ship strikes are rarely recorded in these regions, partly due to the lack of regulations, including no reporting obligation. However, the case studies reveal that another factor may severely augment under-reporting of ship strike mortality. In both instances the crew became aware of the collision only upon arrival at port, suggesting that whales which are hit and killed or wounded, but do not become draped over the bow bulb (if the vessel has one), go unnoticed. The probability of bow draping may be low, and modelling should be attempted. National authorities are encouraged to improve data collection and introduce regulations such as mandatory reporting. These are the first fully documented, fatal whale collisions in Ecuador and West Africa (south of the Canary Islands), and the first struck and killed Bryde’s whale in the Southeast Pacific.

RESUMEN: Documentamos dos inusuales casos de ballenas atrapadas en el bulbo de proa de barcos porta contenedores en el Pacífico Sudeste y en el Atlántico tropical oriental. El 10 de diciembre de 2004, el P&O Nedlloyd Pantanal, un barco carguero de 207m de longitud, colisionó con una ballena de Bryde Balaenoptera edeni en la parte Sur del Golfo de Guayaquil, Ecuador. La ballena estaba fresca y mostraba una extensa zona de la piel con hematomas cuando fue examinada. En un encuentro similar, el 19 de marzo de 1998, el barco porta contenedores OSNA Bruck arribó al puerto de Dakar, Senegal, con el cuerpo de una joven ballena sei B. borealis sobre el bulbo de proa. La colisión ocurrió entre Las Palmas, Gran Canaria y Dakar, Senegal. Balaenopteridos recién muertos son difícilmente recogidos por los barcos debido a que al morir no flotan sino hasta ya bien avanzado el proceso de descomposición, por lo que no habría duda de que en ambos casos las ballenas estaban vivas cuando la colisión ocurrió. Una revisión de los casos de colisiones de barcos con ballenas ocurridos previamente en ambas regiones demuestra que este tipo de eventos son raramente registrados, en parte debido a la falta de regulaciones respecto a la obligatoriedad de reportarlos a las autoridades. Sin embargo, habría otro factor que explica el bajo numero de reportes de ballenas muertas por colisiones; en ambos casos las tripulaciones se percataron de la colisión solo cuando llegaron a puerto, sugiriendo que las ballenas muertas o heridas producto de una colisión que no quedan atrapadas en el bulbo de proa (si es que el barco tiene uno), pasan inadvertidas. La probabilidad que ballenas queden atrapadas en esta parte del barco parecería baja, aunque un estudio de modelación ayudaría a una mejor compresión del problema. Se recomienda a las autoridades nacionales mejorar la colección de información sobre el tema introduciendo disposiciones tendientes a hacer obligatoria su notificación. Estos son los primeros casos bien documentados de colisiones fatales para ballenas en Ecuador y Africa occidental (al sur de las islas Canarias) y la primera ballena de Bryde muerta por esta causa en el Pacífico Sudeste.

KEYWORDS: ship strikes, bow bulb, Bryde’s whale, humpback whale, sei whale, Ecuador, Senegal, West Africa, Balaenoptera edeni, Balaenoptera borealis.

Introduction In general, collisions occur in coastal areas where whales concentrate for feeding or breeding (Laist et al., 2001). The The increase of both maritime traffic and vessel speed is most frequently involved mysticete species include fin cause of concern because of the high number of cetaceans (Balaenoptera physalus), right (Eubalaena glacialis and E. involved in collisions around the world (IWC, 2002; Reeves australis), humpback (Megaptera novaeangliae), gray et al., 2003). In their global compilation, Jensen and Silber (Eschrichtius robustus), common minke (B. acutorostrata) (2004) informed about 292 cases of ship strikes with whales and blue whales (B. musculus). The sperm whale (Physeter between 1975 and 2002, most of them in waters of the macrocephalus) is the most common odontocete involved United States. However, as cautioned by the authors, these (Laist et al., 2001; Jensen and Silber, 2004). The impact on cases would represent a small part of the total cases whale populations is unknown, but in the case of the North occurred since most collisions are unperceived or under- Atlantic right whales (E. glacialis), collisions with vessels reported by crews. Collisions with whales occur with all were responsible for 35.5% of total recorded mortality in types of vessels including cargo, tankers, cruise and fishing the period 1970-1999 (Knowlton and Kraus, 2001). NOAA vessels, although they are more frequent with bigger and Fisheries is currently developing a strategy to reduce faster vessels (Laist et al., 2001; Jensen and Silber, 2004). whale mortality by ship strikes which includes, among

1 Fundación Ecuatoriana para el Estudio de Mamíferos Marinos (FEMM), PO Box 09-01-11905. Guayaquil, Ecuador. E-mail: [email protected]; URL: www.femm.org. 2 Peruvian Centre for Cetacean Research (CEPEC), Museo de Delfines, Pucusana, Lima-20, Peru. E-mail: [email protected]. 090 56 F.FÉLIX AND K. VAN WAEREBEEK others, operational measures for vessels larger than 65ft the sailboat Joel struck a whale, possibly a sperm whale, (19.8m), such as the modification of navigation routes to 120nm offshore during a sailing championship between avoid areas of whale concentration, speed restrictions, and the Ecuador mainland and the Galápagos Islands; and (3) a dynamic management of the area (Silber et al., 2004). in August 2001 an open fishing boat was destroyed when Here we document two recent cases off Ecuador and off it ran into a humpback whale near La Plata Island. In Peru, West Africa, in which two species infrequently reported the NOAA’s R/V Surveyor struck an undetermined whale in ship strikes, the Bryde’s whale (Balaenoptera edeni) and 19km off Callao (Laist et al., 2001; Jensen and Silber, 2004). the sei whale (Balaenoptera borealis) were involved. With A blue whale B. musculus that run aground, bleeding, on the purpose to evaluate their frequency, we reviewed the rocks of isla Don Martín, central Peru, in January 1997 the known cases of collisions with whales in the Southeast may have collided with a ship (Van Waerebeek et al., 1997). Pacific and the Eastern Tropical Atlantic Oceans. Goya et al. (2004) and Luis Santillán (CEPEC, unpublished data) suggested that a ship strike could be the cause of a sperm whale stranded at Paramonga, Peru. Ship strikes in the Southeast Pacific

The frequency of ship strikes with whales in the Southeast Ecuador Pacific is poorly known, even though it is an emerging cause of concern as an unassessed, but potentially At daybreak on 10 December 2004, the 207m-length cargo significant, source of anthropogenic mortality due to the ship P&O Nedlloyd Pantanal, en route from Callao, Peru, important maritime routes across the region (Flórez et al., arrived at the quarantine area of the port of Guayaquil, in press). Most documented collisions in the Southeast Ecuador, with a freshly dead Bryde’s whale draped over Pacific involve humpback whales. According to Capella the bow bulb. The species was identified by the et al. (2001), at least three cases of humpback whales found diagnostic presence of three head ridges, colouration dead between 1986 and 2000 in Colombia were believed pattern and the form and size of the dorsal fin. to be caused by ship strikes. Photos of humpback whales From information provided by the captain, on the night in Ecuador also show wounds on the back, or tails with of 9 December, between 20:00h and 21:00h, the ship’s speed an entire fluke missing, likely caused by propellers (FEMM dropped from 18.5 to 16.8kn without obvious reason. catalogue3, unpublished data). Haase and Félix (1994) Engine power had to be increased to reach the entrance of informed of a 12.6m sperm whale with fractured the inner estuary of the Gulf of Guayaquil on schedule. maxillaries beached in 1991 at Punta Carnero, Ecuador, At the moment when speed decreased, and presumably probably caused by a ship strike. In FEMM’s database the collision occurred, the ship was in the southern part there are also three other anecdotal cases of ship strikes of the Gulf (03°34’S,80°58’W - 03.20°S, 80°48’W), at the published in Ecuadorian newspapers: (1) in August 1989 border between Ecuador and Peru (Figure 1). The impact the small purse-seiner Paquín sunk after collision with a was not perceived by the crew who realised the event only whale, presumably a humpback whale; (2) in October 1996 when the ship entered port.

Figure 1. Area (hatched circle) at the entrance of the Gulf of Guayaquil where the collision with a Bryde’s whale occurred on 9 December 2004, according to information received from the captain of the cargo ship P&O Nedlloyd Pantanal.

3 Currently the FEMM catalogue contains photographic evidence of individual identification for around 500 different individuals.

LAJAM 4(1): 55-60, January/June 2005

091 WHALE MORTALITY FROM SHIP STRIKES IN ECUADOR AND WEST AFRICA 57

The whale was positioned on the bow bulb on its Ship strikes in the Eastern Tropical Atlantic belly, midbody out of the water (Figure 2). The area of main impact, however, was not visible because the The oldest report is of a sperm whale that was struck bow’s sharp edge was embedded in the left flank, so halfway between Dakar and Cap Verde Islands on 9 the full extent of trauma could not be evaluated. The June 1955 which seriously damaged the bow of the specimen was freshly dead with most of the skin intact vessel (Cadenat, 1956); it is unclear what happened to and retaining pigmentation: dark gray dorsally, light the whale. The only other area off West Africa for which gray on the flanks and a white throat. On its right data are available is the Canary Islands. Tregenza et al. side, over an area of ca. 4m between the flipper and (2002) compiled 21 probable instances of lethal collisions the dorsal fin, the epidermis was abraded. The in six species of cetaceans between 1985 and February exposed blubber showed massive dermal hematoma, 2002. Since the introduction of fast ferries in 1999, a from behind the ventral grooves rearward, beyond significant increase of fatal collisions occurred, the dorsal fin. A smaller area of damaged epidermis especially with sperm whales. and bruises extended from the ear region forward On 19 March 1998, a dead juvenile sei whale was through the right side of the head. Flippers and the brought into quay no.2 of the port of Dakar, Senegal, dorsal fin were complete, although most of the draped over the bow bulb of the German container ship epidermis of the flippers was ripped off. OSNA Bruck (Group Somico-Smith and Kraft).4 Its body The extensive hematoma evident on the dermis length was about 12m, and its body weight 9,960kg, as indicated that the whale was alive when struck. The left- weighed at Dakar port facilities (Figure 2). The absence side impact, it is thought, may have broken ribs and of bloating, all baleen plates attached to the palate and ruptured vital inner organs, causing massive largely intact skin suggested that the whale had not hemorrhage and the death of the whale. When hit by been dead for more than two days when photographed. the bow bulb, and before being wedged between it and Although the collision was not registered by the crew the sharp bow, the whale may have rolled around its and there is no direct evidence that the whale was killed axis, which would explain skin abrasion also on its right by the impact, the very fresh state of the carcass virtually side. The specimen was an adult-sized female. No excludes that the animal was scooped up dead. Sei standard length could be taken, but FF measured 16.2m whales, like all balaenopterids, sink immediately after along the body’s bended contour. The carcass was death and rise to the surface only after decomposition dropped off the vessel in open waters two days later. and bloating sets in (Fraser, 1937; Slijper, 1979).

Figure 2. Adult Bryde’s whale draped over the bow bulb of container vessel P&O Nedlloyd Pantanal, in Guayaquil port, Ecuador. Note extensive dermal hematoma on the lower mid-body, indicating antemortem trauma.

4 Information was collected by Dr. Papa Ndiaye, Laboratoire Biologie Marine, Institut Fondamental d’Afrique Noire (IFAN), Dakar, as part of the UNEP/CMS WAFCET-1 Project (Van Waerebeek et al., 2000).

LAJAM 4(1): 55-60, January/June 2005

092 58 F.FÉLIX AND K. VAN WAEREBEEK

Figure 3. Young sei whale brought into Dakar port, Senegal, on 23 March 1998, on the bow of the German container ship OSNA Bruck. A full set of baleen still in the palate, mostly intact skin and the lack of bloating suggested that the whale had died within the past two days, presumably from impact with the vessel. Photos by Dr P. Ndiaye, IFAN (CMS/UNEP WAFCET-1 Project).

On Monday 23 March 1998, the Dakar daily Le Soleil whale had found its end north of the port´. No dates published a photo of the whale draped over the ship’s were available for these cases. Ndiaye thought these bow bulb and reported that it was struck close to Gorée were also sei whales, but without voucher samples their Island, a few nautical miles off Dakar. However, specific identity cannot be confirmed. according to the captain of the OSNA Bruck, the ship had departed Las Palmas, Gran Canaria, Canary Discussion Islands, some two days earlier and the crew became aware of the dead whale only after passing Gorée Island. The collision of the container ship P&O Nedlloyd Pantanal The captain indicated that the collision may have with a Bryde’s whale is the first formally recorded case of occurred earlier, en route from Las Palmas. this type of incidental mortality in Ecuador and the first Dr. Papa Ndiaye photographed and collected the time this species is recorded killed by a ship strike in the specimen, however no necropsy was performed. The Southeast Pacific. However, similar cases probably have carcass was trucked to Sangalkam and buried, occurred in open waters, were not noticed or reported, so destined for future retrieval of the skeleton5. No the magnitude of ship collisions with cetaceans is samples were taken except for two apical baleen unassessed in this region. Crews of large cargo vessels plates, deposited at IFAN collection, which were all- generally are unaware of collisions (Jensen and Silber, black with very fine, white bristles (examined by 2004; this paper) and typically notice the kill only when KVW). The arched rostrum with downturned tip, a the whale becomes stuck on the bow. Another reason for single central but no auxiliary rostral ridges, overall limited number of records would be the lack of regulations dark grey colouration, a high falcate dorsal fin and (or enforcement) regarding ship strikes with whales, the baleen were diagnostic for B. borealis. One flipper including no reporting obligation in countries bordering measured 156cm. the Southeast Pacific. In Peru, proposals6 for the Port Captain Mr. Edouard Sarr claimed that this was construction of a mega port (Hub Port) at Isla San Lorenzo- the third whale found in Dakar waters ‘over a short Callao to accommodate ULCS (Ultra Large Container period’. One had stranded at poste 17, and a second Ships), SPPS (Super Post Pamamax Ship) and SCS (Super

5 The skeleton was abandoned since its burial site was developed for urbanization (P. Ndiaye, pers.comm. to KVW). 6 See ‘Proyecto Ciclópeo’in Peruvian weekly CARETAS Edición 1860, 10 February 2005. www.caretas.com.pe.

LAJAM 4(1): 55-60, January/June 2005

093 WHALE MORTALITY FROM SHIP STRIKES IN ECUADOR AND WEST AFRICA 59

Cruiser Ships) greater than 300,000 MT, and additional strike is difficult, but the careful analysis of whale projects for major port expansion at Bayovar, Eten, San carcasses will provide an important line of evidence to Juan and Ilo, are an increasing cause of concern for both estimate strike frequencies and the species involved. It resident and migrating whales off Peru’s coast. would be desirable that environmental authorities in The situation is very similar in the Eastern Tropical coastal nations issue regulations to deal with beached Atlantic. The ship strike with a sei whale, so far we whales and systematically conduct necropsies. Moreover, know, is the first fully documented case off West Africa. national legislation should be reviewed as to include However, high potential for shipping-caused mortality obligatory reporting by ships. In June 2005, the exists in the northern Gulf of Guinea, and specifically Conservation Committee of the International Whaling in the Bight of Benin. Expanding merchant shipping to Commission (IWC) recognised the relevance of these and from Cotonou, Benin, and shipping lanes which issues and instated an ad hoc Ship Strikes Working Group cross the breeding area of a northern Gulf of Guinea (SSWG) to examine scientific, technical and policy humpback whale stock (Van Waerebeek et al., 2001, aspects, and where necessary to recommend action. It 2002) pose increasing risks and endangers slow-moving was agreed that special attention be paid to critical areas mother/calf pairs in particular. A second container port where high density shipping channels are over-laid with is planned along Benin’s coast and heavy shipping known whale aggregation areas (IWC, 2005)8. traffic is linked to Nigerian (Lagos, Port Harcourt), Uncertainty exists regarding the population identity Ghanaian (Tema, Sekondi-Takoradi) and Cameroon and status of sei whales off West Africa. On geographic (Douala) ports. Under current circumstances ship grounds, they would tentatively form part of the collisions with cetaceans off West Africa may often Eastern North Atlantic stock (sensu Jonsgård and remain formally unreported, and mortality Darling, 1977; Donovan, 1991). Allen (1916) claimed that unaccounted for. National authorities are encouraged sei whales were rarely seen south of the Straits of to improve data collection and introduce regulations Gibraltar, contradicted by Anonymus (1914, in Jonsgård including obligatory reporting. and Darling, 1977) who indicated that sei whales were Ship strikes with both Bryde’s and sei whales are observed, sometimes in large numbers, in the area that considered uncommon. The NOAA database contains stretches southward from the Madeira Islands, past the only three cases for each of these species world-wide Canary toward the Cape Verde Islands. Ingebrigtsen (Jensen and Silber, 2004). Unpublished reports exist of (1929) argued that sei whales stay in the southern part Bryde’s whales struck and killed by ships travelling to of the North Atlantic during the winter and their or from Auckland’s busy commercial port, in the northward migrations seem to take place offshore. Hauraki Gulf, New Zealand (Alan N. Baker, pers. Kirpichnikov (1950) reported on three sightings in May comm.). Perhaps part of the explanation is that these 1948 off West Africa, the southernmost record being at species are among the fastest moving baleen whales 06º30’N, 18º20’W. However, Jonsgård (1966) warned (Slijper, 1979). Their smaller size, compared to other that in this area Bryde’s whales sighted may easily be more frequently affected species, such as fin and right confused with sei whales. Maigret (1981) reported on a whales, also reduces the probability to be struck. stranding of an 11m B. borealis in February 1981 in the Vessel speed seems the most relevant factor driving ship Baie du Lévrier, Mauritania. In Senegal a series of nine strikes: 90% of cases in which the speed was known, baleen plates diagnostic for a sei whale were recovered vessels moved at 10kn or higher, with the highest rate from a tiger shark stomach landed at Joal in July 1949 of incidence between 13 and 18kn (Jensen and Silber, (Cadenat, 1955). The present specimen is the first 2004). The cruise speed of the P&O Nedlloyd Pantanal is documented case in Senegal for half a century and even higher. Considering the ‘hull speed’7 of a vessel is apparently only the third authenticated specimen directly proportional to the square root of its waterline record for West Africa. No evidence of sei whales has length, collision frequency must also be a function of been found in The Gambia and Guinea-Bissau (Van ship size. Modelling of ship strike dynamics may offer Waerebeek et al., 2000). valuable insights. A large proportion of ship strike records in the NOAA Acknowledgements database are from carcasses of beached animals, which show signs of cuts by propellers, as well as fractures of FF thanks the Ecuador Environment and Port skull and ribs (Jensen and Silber, 2004). According to Laist Authorities for the facilities provided to examine the et al. (2001), 58 of 407 beached whales (14%) recorded in Bryde’s whale in Guayaquil; the captain of the P&O the USA east coast between 1975 and 1996 could have Nedlloyd Pantanal, Mr. Rudolph Gruhnwald, who been caused by ship strikes. Diagnosing death from ship kindly provided information. Dr. Papa Ndiaye and Mr.

7 The maximum speed a given hull can attain in a displacement vessel (i.e. not planing vessels). 8 International Whaling Commission (2005) Report of the Conservation Committee. IWC/57/Rep 5, IWC Annual Meeting, Ulsan, Korea, May-June 2005.

LAJAM 4(1): 55-60, January/June 2005

094 60 F.FÉLIX AND K. VAN WAEREBEEK

Abdoulaye Djiba of IFAN (Dakar) are thanked for Dolphins and Porpoises. University of California Press, Berkeley and logistics support during survey work in Senegal. Dr. Los Angeles. Nick Gales and Dr. A.N. Baker kindly provided JONSGÅRD, A. AND DARLING, K. (1977) On the biology of the eastern reviews with several helpful comments on very short North Atlantic sei whale, Balaenoptera borealis Lesson. Report notice. The WAFCET-1 Project, directed by KVW, was International Whaling Commission (special issue 1): 124-129. sponsored by the Convention on the Conservation of KIRPICHNIKOV, A.A. (1950) Nablyudeniya nad raspredeleniyem Migratory Species of Wild Animals (CMS/UNEP), kito-obraznakh v Atlanticheskom okeane. Priroda (Leningrad), Bonn. KVW represented Belgium’s Federal Public 10: 63-64.

Service (Public Health, Food Chain Security and KNOWLTON A. R. AND KRAUS, S.D. (2001) Mortality and serious injury Environment) at the IWC 57th Annual Meeting. The of northern right whales (Eubalaena glacialis) in the Western North International Whaling Commission supported his Atlantic Ocean. Journal of Cetacean Research and Management travel between Lima and Brussels. (special issue 2): 193-208. LAIST, D. W., KNOWLTON, A.R., MEAD, J.G., COLLET, A.S. AND PODESTA, REFERENCES M. (2001) Collision between ships and whales. Marine Mammal Science 17(1): 35-75.

ALLEN, G.M. (1916) The whalebone whales of New England. MAIGRET, J. (1981) Les mammifères des côtes de Mauritanie 2. Memoirs of the Boston Society for Natural History 8: 106-322. Rapport annuel des observations signalées en 1981. Bulletin du Centre National de Recherches Océanographiques et des Pêches, ANONYMUS. (1914) Seihvalen. Norsk Hvalfangst Tidende. 43(4): 57-58. Nouadhibou 10(1): 81-85. CADENAT, J. (1955) A propos d’un échouage de baleine à Dakar. REEVES, R. R., SMITH, B.D., CRESPO, E. AND NOTARBARTOLO DI SCIARA, Notes Africaines 67: 91-94. G. (compilers). (2003) Dolphins, Whales and Porpoises: 2002–2010 CADENAT, J. (1956) A propos de cachalot. Notes Africaines 71: 82-92. Conservation Action Plan for the World’s Cetaceans. IUCN/SSC Cetacean Specialist Group. IUCN, Gland, Switzerland and CAPELLA, J., FLÓREZ-GONZÁLEZ, L. AND FALK, P. (2001) Mortality and Cambridge, UK. 139pp. anthropogenic harassment of humpback whales along the Pacific coast of Colombia. Memoirs of the Queensland Museum 47(2): 547-553. SILBER, G., GERRIOR, P. AND ZOODSMA, B. (2004) NOAA’s Fisheries proposed strategy to reduce ship strikes of North Atlantic right DONOVAN, G.P. (1991) A review of IWC stock boundaries. Pages 39- whales – Shipping Industry dialog. Available at: http:// 68 in HOELZEL, A.R. (Ed.) Genetic ecology of whales and dolphins. 311pp. www.nero.noaa.gov/shipstrike. FLÓREZ-GONZÁLEZ, L., CAPELLA, J., FALK, P., FÉLIX, F., GIBBONS, J., HAASE, SLIJPER, E.J. (1979) Whales. Hutchinson of London (second English B., SANTILLÁN, L., PEÑA, V., ÁVILA, I., HERRERA, J., TOBÓN, I. AND VAN edition). 511pp. WAEREBEEK, K. (In press) Estrategia para la Conservación de la Ballena Jorobada del Pacífico Sudeste. Universidad de Magallanes, Chile. TREGENZA, N., AGUILAR, N., CARRILLO, M, DELGADO, I., AND DÍAZ, F. (2002) Collisions between fast ferries and whales in the Canary FRASER, F.C. (1937) Giant Fishes, Whales and Dolphins. Putnam, London. 361pp. Islands: observational data and theoretical limits. Document SC/ 54/BC4 presented to IWC Scientific Committee Meeting. 7pp. GOYA, E., MÁRQUEZ, J.C. AND GARCÍA-GODOS, A. (2004) Informe (unpublished). nacional del Perú sobre el estado actual de los mamíferos marinos y las medidas de protección adoptadas. Informe presentado a la VAN WAEREBEEK, K., PASTENE, L.A., ALFARO-SHIGUETO, J., BRITO, J.L. AND III Reunión de Expertos para Revisar las Actividades del Plan de MORA-PINTO, D. (1997) The status of the blue whale Balaenoptera musculus Acción para la Conservación de los Mamíferos Marinos del off the west coast of South America. Paper SC/49/SH9 presented to Pacífico Sudeste. Lima, Perú, 23-25 de marzo de 2004. 55pp. IWC Scientific Committee Meeting, Bournemouth, UK. 12pp. (Unpublished). VAN WAEREBEEK, K., NDIAYE, E., DJIBA, A., DIALLO, M., MURPHY, P., HAASE, B. AND FÉLIX, F. (1994) A note on the incidental mortality of JALLOW, A., CAMARA, A., NDIAYE, P. AND TOUS, P. (2000) A survey sperm whales (Physeter macrocephalus) in Ecuador. Report of the of the conservation status of cetaceans in Senegal, The Gambia International Whaling Commission (special issue 15): 481-483. and Guinea-Bissau. WAFCET-1 Report. UNEP/CMS Secretariat, Bonn, Germany. 80pp. IWC (2002) Report of the Scientific Committee. Annex M. Estimation of bycatch and other human-induced mortality. Pp.19- VAN WAEREBEEK, K., NOBIMÉ, G., SOHOU, Z., TCHIBOZO, S., DOSSOU- 24. IWC/54/4. BODJRENOU, J.S., DOSSOU, C., AND DOSSOU-HOUNTOUDOU, A. (2002) Introducing whale and dolphin watching to Benin, 2002 INGEBRIGTSEN, A. (1929) Whales caught in the North Atlantic exploratory survey. Report to the Netherlands Committee for and other seas. Rapports et Procès-verbaux des Réunions du IUCN, Amsterdam. 9pp. (unpublished). Conseil Permanent International pour l’Exploration de la Mer 56(1): 1-123. VAN WAEREBEEK, K., TCHIBOZO, S., MONTCHO, J., NOBIME, G., SOHOU, Z., SOHOUHOUE, P. AND DOSSOU, C. (2001) The Bight of Benin, a JENSEN, A. S. AND SILBER, G.K. (2004) Large whale ship strike database. North Atlantic breeding ground of a Southern Hemisphere NOAA Technical Memorandum NMFS-OPR. January 2004. 37pp. humpback whale population, likely related to Gabon and Angola JONSGÅRD, A. (1966) The distribution of Balaenopteridae in the substocks. Paper SC/53/IA21 presented to the IWC Scientific North Atlantic Ocean. Pages 114-124 in NORRIS, K.S. (Ed.) Whales, Committee Meeting, London, July 2001. 8pp.

Received 28 April 2005. Accepted 15 June 2005.

LAJAM 4(1): 55-60, January/June 2005

095 www.jpsr.org Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013

Modeling Oil and Petroleum Evaporation Merv F. Fingas Spill Science 1717 Rutherford Point, S.W., Edmonton, AB, Canada T6W 1J6 *[email protected]

Abstract prime role in the fate of most oils. In a few days, Evaporation is an important component in oil spill models. typical crude oils can lose up to 45% of their volume Various approaches for oil evaporation prediction are (Fingas 2011). The Deepwater Horizon oil lost up to summarized. Models can be divided into those models that 55% in a short time when released under water at high use the basis of air‐boundary‐regulation or those that use pressure. Many crude oils must undergo evaporation liquid diffusion‐regulated evaporation physics. Studies before the formation of water‐in‐oil emulsions. Light show that oil is not air boundary‐layer regulated such as it is oils will change very dramatically from fluid to for water evaporation, which implies that a simplistic viscous; while heavy oils will become solid‐like. Many evaporation equation suffices to accurately describe the oils after long evaporative exposure, form tar balls or process. The following processes do not require heavy tar mats. Despite the importance of the process, consideration: wind velocity, turbulence level, area and scale size. The factors important to evaporation are time and only some work has been conducted on the basic temperature. Oil evaporation does show a thickness effect, physics and chemistry of oil spill evaporation (Fingas although not as pronounced as that for air‐boundary‐layer 1995). The difficulty in studying oil evaporation is that regulated models. A thickness adjustment calculation is oil is a mixture of hundreds of compounds and oil presented for diffusion‐regulated models. This new model is composition varies from source to source and even applicable to thicknesses greater than about 1.5 mm. In the over time. Much of the work described in the previous case of thin slicks, this adjustment is not relevant as oils literature focused on calibrating equations developed typically spread to less than that in a short time. for water evaporation (Fingas 1995). The use of air‐boundary‐models results in three types of The mechanisms that regulate evaporation are errors: air‐boundary‐layer models cannot accurately deal with long term evaporation; second, the wind factor results important (Brutsaert 1995; Jones 1992). Evaporation of in unrealistic values and finally, they have not been adjusted a liquid can be considered as the movement of for the different curvature for diesel‐like evaporation. molecules from the surface into the vapour phase Further, these semi‐empirical equations require inputs such above it. The immediate layer of air above the as area, etc., that are unknown at the time of the spills. There evaporation surface is known as the air boundary has been some effort on the part of modellers to adjust air‐ layer5 which is the intermediate interface between the boundary‐layer models to be more realistic on the long‐term, air and the liquid and might be viewed as very thin e.g. but these may be artificial and result in other errors such as as less than 1 mm. The characteristics of this air under‐estimation for long‐term prediction. A comparison of boundary layer can influence evaporation. In the case models shows that on a very short term, such as a few hours, most models yield similar results. However, as time of water, the boundary layer regulates the evaporation increases past a few days, the errors with air‐boundary‐layer rate. Air can hold a variable amount of water, regulated models are unacceptable. Examples are given depending on temperature, as expressed by the where errors are as large as 100% over a few days. relative humidity. Under conditions where the air Keywords boundary layer doesn’t move (no wind) or has low turbulence, the air immediately above the water Oil Spill Evaporation; Hydrocarbon Evaporation; Evaporation quickly becomes saturated and evaporation slows. The Modeling actual evaporation of water proceeds at a small fraction of the possible evaporation rate because of the Introduction saturation of the boundary layer. The air‐boundary‐ Evaporation is an important process for most oil spills. layer physics is then said to regulate the evaporation Almost all oil spill models include evaporation as a of water. This regulation manifests the increase of process and output of the model. Evaporation plays a evaporation with wind or turbulence. When

104 096 Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013 www.jpsr.org turbulence is weak, evaporation can slow down by high probability that the air is saturated downwind orders‐of‐magnitude. The molecular diffusion of water and the evaporation rate per unit area is lower than molecules through air is at least 103 times slower than that for a smaller pool. It is noted that there are many turbulent diffusion (Monteith and Unsworth 2008). equivalent ways to express this fundamental evaporation equation. These will be seen in the Some liquids are not air‐boundary‐layer regulated equations below. primarily because they evaporate too slowly to make the vapours saturate the air boundary layer above Sutton proposed the following equation based on them (Fingas 2011). Many mixtures are regulated by empirical work (Brutsaert 1982): the diffusion of molecules inside the liquid to the E = KC U7/9 d -1/9 Sc- r (2) surface of the liquid. Such a mechanism is true for s many slowly‐evaporating mixtures of compounds where Cs is the concentration of the evaporating fluid such as oils and fuels. Some of the outcomes of this (mass/volume), U is the wind speed, d is the area of mechanism may seem counterintuitive to some people the pool, Sc is the Schmidt number and r is the such as that increasing area may not necessarily empirical exponent assigned values from 0 to 2/3. increase evaporation rate. More importantly, Other parameters are defined as above. The terms in increasing wind speed does not increase evaporation. this equation are analogous to the very generic equation, (1), proposed above. The turbulence is Scientific work on water evaporation dates back expressed by a combination of the wind speed, U, and decades and thus the basis for early oil evaporation the Schmidt number, Sc that is the ratio of kinematic work has been established (Fingas 2011). There are viscosity of air (ν) to the molecular diffusivity (D) of several fundamental differences between the the diffusing gas in air, i.e., a dimensionless expression evaporation of a pure liquid such as water and that of of the molecular diffusivity of the evaporating a multi‐component system such as crude oil. The substance in air.7 The coefficient of the wind power evaporation rate for a single liquid such as water is a typifies the turbulence level. The value of 0.78 (7/9) as constant with respect to time. Evaporative loss, either chosen by Sutton, represents a turbulent wind by weight or volume, is not linear with time for crude whereas a coefficient of 0.5 would represent a wind oils, and other multi‐component fuel mixtures (Fingas flow that is more laminar. The scale length 1997). represented by d has been given an empirical exponent of ‐1/9. This represents for water, a weak Review of Historical Developments dependence on size. The exponent of the Schmidt For air‐boundary‐layer regulated liquids, one can number, r, represents the effect of the diffusivity of the write the mass transfer rate in semi‐empirical form as particular chemical, and historically was assigned (Fingas 2011): values between 0 and 2/3 (Sutton 1934).

E = K C Tu S (1) Blokker was the first to develop oil evaporation equations for oil evaporation at sea, with his partially where E is the evaporation rate in mass per unit area, theoretical starting basis (Blokker, 1964). Oil was K is the mass transfer rate of the evaporating liquid, presumed to be a one‐component liquid. The sometimes denoted as kg (gas phase mass transfer distillation data and the average boiling points of coefficient, which may incorporate some of other successive fractions were used as the starting point to parameters noted here), C is the concentration (mass) predict an overall vapour pressure. The average of the evaporating fluid as a mass per volume, Tu is a vapour pressure of these fractions was then calculated factor characterizing the relative intensity of from the Clausius‐Clapeyron equation to yield: turbulence, and S is a factor related to the saturation of p 1 1 the boundary layer above the evaporating liquid. The log s = qM - (3) p 4.57 T T saturation parameter, S, represents the effects of local s advection on saturation dynamics. If the air has where p is the vapour pressure at the absolute tem‐ already been saturated with the compound in question, perature, T; ps is the vapour pressure at the boiling the evaporation rate approaches zero. This also relates point, Ts (for ps, 760 mm Hg was used); q is the heat of to the scale length of an evaporating pool. If one views evaporation in cal/g and M is the molecular weight. a large pool over which a wind is blowing, there is a The term qM/(4.57 Ts) was taken to be nearly constant

105 097 www.jpsr.org Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013 for hydrocarbons (=5.0 +/‐ 0.2) and thus the expression sea Butler (1976) developed a model to examine was simplified to evaporation of specific hydrocarbon components. The weathering rate was taken as proportional to the log ps /p = 5.0 [ (Ts ‐ T)/T] (4) equilibrium vapour pressure, P, of the compound and From the empirical data and equation (4), the to the fraction remaining: weathering curve was calculated, assuming that dx/dt = ‐kP(x/xo) (7) Raoultʹs law is valid for this situation giving qM as a function of the percentage evaporated. Pasquillʹs where x is the amount of a particular component of a equation was applied stepwise, and the total crude oil at time, t, xo is the amount of that same com‐ evaporation time was obtained by summation: ponent present at the beginning of weathering (t = 0), k is an empirical rate coefficient and P is the vapour pressure of the chosen oil component. where t is the total evaporation time in hours, Δh is the Butler assumed that petroleum is a complicated decrease in layer thickness in m, D is the diameter of mixture of compounds, therefore P is not equal to the the oil spill, β is a meteorological constant (assigned a vapour pressure of the pure compound, but neither value of 0.11), Kev is a constant for atmospheric would there be large variation in the activity stability (taken to be 1.2 x 10‐8), α is a meteorological coefficient as the weathering process occurs (Butler constant (assigned a value of 0.78), P is the vapour 1976). For this reason, the activity coefficients were pressure at the absolute temperature, T; and M is the subsumed in the empirical rate coefficient k. P and k molecular weight of the component or oil mass. Tests were taken as independent of the amount, x, for a of this equation by experimental evaporation using a fairly wide range of oils. The equation was then small wind tunnel did not yield good correspondence directly integrated to give the fraction of the original to test data. compound remaining after weathering as:

Mackay and Matsugu (1973) approached evaporation x/xo = exp(‐ktP/xo) (8) by using the classical water evaporation and experimental work. The water evaporation equation The vapour pressure of individual components was fit was corrected to hydrocarbons using the evaporation using a regression line to yield a predictor equation for rate of cumene. Data on the evaporation of water and vapour pressure: cumene have been used to correlate the gas phase P = exp(10.94 ‐ 1.06 N) (9) mass transfer coefficient as a function of wind‐speed and pool size by the equation, where P is the vapour pressure in Torr and N is the carbon number of the compound in question. This 0.78 ‐0.11 ‐0.67 K m = 0.0292 U X Sc (6) combined with equation (8) and yielded the following

Where Km is the mass transfer coefficient in units of expression: mass per unit time and X is the pool diameter or the x/xo = exp [‐(kt/xo)exp(10.94 ‐ 1.06 N)] (10) scale size of evaporating area. Note that the exponent of the wind speed, U, is 0.78 equal to the classical Where x/xo is the fraction of the component left after water evaporation‐derived coefficient. Mackay and weathering, k is an empirical constant, xo is the Matsugu noted that for hydrocarbon mixtures the original quantity of the component and N is the evaporation process is more complex, dependent on carbon number of the component in question. the liquid diffusion resistance being present.9 Equation (10) predicts that the fraction weathered is a Experimental data on gasoline evaporation were function of the carbon number and decreases at a rate compared with computed rates which showed some that is faster than predicted from simple exponential deviations from the experimental values and decay.12 If the initial distribution of compounds is suggested the presence of a liquid‐phase mass‐transfer essentially uniform (xo independent of N), then the resistance. The same group showed that the above equation predicts that the carbon number where evaporative loss of a mass of oil spilled can be a constant fraction (e.g. half) of the initial amount has estimated using a mass transfer coefficient, Km, as been lost (x = 0.5 xo) is a logarithmic function of the shown above (Goodwin et al. 1976). This approach time of weathering: was investigated with some laboratory data and tested N1/2 = 10.66 + 2.17 log (kt/xo) (11) against some known mass transfer conditions on the

106 098 Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013 www.jpsr.org

where N1/2 is half of the volume fraction of the oil. The that the equations yielded predictions that were close equation was tested using evaporation data from some to the experimental data. Rheijnhart and Rose (1982) patches of oil on shoreline, whose age was known. The developed a simple predictor model for the equation was capable of predicting the age of the evaporation of oil at sea and proposed the following samples relatively well. It was suggested that the simple relationship: equation was applicable to open water spills; however, Qei = αCo (16) this was never subsequently applied in models. where Qei is the evaporation rate of the component of Yang and Wang (1977) developed an equation using interest, α is a constant incorporating wind velocity the Mackay and Matsugu molecular diffusion and other factors (taken as 0.0009 m s‐1) and Co is the process.2 The vapour phase mass transfer process was equilibrium concentration of the vapour at the oil expressed as: surface. Several pan experiments were run to simulate evaporation at sea and the data used to test the equation. No method was given to calculate the where Die is the vapour phase mass transfer rate, km is essential value, Co. a coefficient that lumps all the unknown factors Brighton (1985,1990) proposed that the standard affecting the value of Die, pi is the hydrocarbon vapour formulation used by many workers required refining. pressure of fraction, I, at the interface, pi∞ is the His starting point for water evaporation was similar to hydrocarbon vapour pressure of fraction, I, at infinite that proposed by Sutton: altitude of the atmosphere, R is the universal gas 7/9 1/9 - r constant and Ts is the absolute temperature of the oil E = K C U d Sc (17) m s slick. The following functional relationship was where E is the mean evaporation rate per unit area, Km proposed (Yang and Wang 1977): is an empirically‐determined constant, presumably related to the foregoing mass transfer constant, Cs is where A is the slick area, U is the over‐water wind the concentration of the evaporation fluid speed, and a, q and γ are empirical coefficients. This (mass/volume), d is the area of the pool and r is an relationship was based on the results of previous empirical exponent assigned values from 0 to 2/3. studies, including, for instance, those of MacKay and Brighton suggested that this equation should conform Matsugu who suggested the value of γ to be in the to the basic dimensionless form involving the range from ‐0.025 to ‐0.055.9 Further experiments were parameters U and Zo (wind speed and roughness performed by Yang and Wang to determine the values length, respectively) which define the boundary layer of ‘a’ and ‘q’. Experiments showed that a film formed conditions. The key factor in Brighton’s analysis was on evaporating oils and this film severely retarded to use a linear eddy‐diffusivity profile. This feature evaporation. Before the surface film has developed implied that concentration profiles become logarithmic near the surface, which is suspected to be (ρt/ρo < 1.0078): more realistic compared to the more finite values previously used. Using a power profile to provide an where Kmb is the coefficient that groups all factors estimation of the turbulence, Brighton was able to affecting evaporation before the surface film has substitute the following identities into the classical formed and A is the area. After the surface film has relationship: developed (ρt/ρo > 1.0078)

Kma = 1/5 kmb (15) where ρo is initial oil density, ρt is weathered oil density at time t, and Kma is the coefficient that groups Where: u* is the friction velocity, z1 is the reference all factors affecting evaporation after the surface film height above the surface, z0 is the roughness length 12 has formed. The evaporation rate was found to be and n is the power law dimensionless term. The reduced fivefold after the formation of the surface film. evaporation equation now became: Drivas (1982) compared the Mackay and Matsugu equation with data found in the literature and noted

107 099 www.jpsr.org Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013

where z is the height above the surface, Χ is the dFv = [Pν/(RT)](KAdt/Vo) (27) concentration of the evaporating compounds, x is the or dFv = Hdθ (28) dimension of the evaporating pool, k given by K/u*z, is the von Karman constant and σ is the turbulent where H is Henryʹs law constant and θ is the Schmidt number (taken as 0.85). Brighton evaporative exposure (defined below). subsequently compared his model with experimental The right‐hand side of the second last equation has evaporation data in the field and in the laboratory, been separated into two dimensionless groups (Stiver including laboratory oil evaporation data (Brighton and MacKay 1984). The group, KAdt/Vo, represents 1985, 1990). The model only correlated well with the time‐rate of what has been termed as the laboratory water evaporation data and the reason “evaporative exposure” and was denoted as dθ. The given was other data sets were ‘noisy’. evaporative exposure is a function of time, the spill Tkalin (1986) proposed a series of equations to predict area and volume (or thickness), and the mass transfer evaporation at sea: coefficient (which is dependent on the wind speed). The evaporative exposure can be viewed as the ratio of K M P x = a i oi t (21) exposed vapour volume to the initial liquid volume.18 Ei RT The group Pν/(RT) or H is a dimensionless where Ei is the evaporation rate of component I (or the Henryʹs law constant or ratio of the equilibrium 2 sum of all components) (kg/m s), Ka is the mass concentration of the substance in the vapour phase transfer coefficient (m/s), Mi is the molecular weight, [P/(RT)] to that in the liquid (l/ν). H is a function of Poi is the vapour pressure of the component I, and xt is temperature. The product θH is thus the ratio of the the amount of component I at time, t. Using empirical amount which has evaporated (oil concentration in data, relationships were developed for some of the vapour times vapour volume) to the amount originally factors in the equation: present. For a pure liquid, H is independent of Fv and Poi = 103eA (22) equation 26 was integrated directly to give: where A = ‐(4.4 + logTb)[1.803{Tb/T ‐ 1} ‐ 0.803 ln(Tb/T)] Fv = H θ (29) (23) If K, A, and temperature are constant, the evaporation and where Tb is the boiling point of the hydrocarbon, rate is constant and evaporation is complete (Fv is given as unity) when θ achieves a value of 1/H.

‐3 Ka = 1.25U10 (24) If the liquid is a mixture, H depends on Fv and the The equations were verified using empirical data from basic equation can only be integrated if H is expressed the literature. as a function of Fv; i.e., the principal variable of vapour pressure is expressed as a function of composition. A frequently used work in older spill modelling is that The evaporation rate slows as evaporation proceeds in of Stiver and Mackay (1984)based on some of the such cases. Equation (27) was replaced with a new earlier work of Mackay and Matsugu (1973). The equation developed using laboratory empirical data: formulation was initiated with assumptions on the evaporation of a liquid. If a liquid is spilled, the rate of Fv = (T/K1) ln (1 + K1θ/T) exp(K2 ‐ K3/T) (30) evaporation is given as: where Fv is the volume fraction evaporated and K1,2,3 N = KAP/(RT) (25) are empirical constants.18 A value for K1 was obtained from the slope of the Fv vs. log θ curve from pan or where N is the evaporative molar flux (mol/s), K is the bubble evaporation experiments. For θ greater than mass transfer coefficient under the prevailing wind 104, K1 was found to be approximately 2.3T divided by (ms‐1) and A is the area (m2), P is the vapour pressure the slope. The expression exp(K2 ‐ K3/T) was then of the bulk liquid. This equation was arranged to give: calculated, and K2 and K3 were determined dFv/dt = KAPν/(VoRT) (26) individually from evaporation curves at two different temperatures. where Fv is the volume fraction evaporated, v is the liquidʹs molar volume (m3/mol) and Vo is the initial Hamoda and co‐workers (1989) performed theoretical volume of spilled liquid (m3). By rearranging: and experimental work on evaporation. An equation

108 100 Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013 www.jpsr.org was developed to express the effects of APIo In summary, it is difficult to develop a theoretical (American Petroleum Institute gravity‐a unit of approach to oil evaporation for several reasons. First, density) of the crude oil, temperature, and salinity on oil consists of many components and thus there is no the mass transfer coefficient K: constant boiling point, vapor pressure or other essential properties used in typical evaporation K = 1.68 x 10‐5 (APIo)1.253 (T)1.80 e0.1441 (31) models. Further, oil evaporation proceeds by diffusion where K is the mass transfer coefficient, cm h‐1, APIo is regulation but not by air‐boundary‐layer regulation. the density in API units, unitless, and e is the water Water evaporation models cannot be accurately salinity in degrees salinity or parts‐per‐thousand. The modified to oil evaporation for these reasons. exponents of the equation were determined by multiple linear regression on experimental data. Development of Diffusion-Rgulated Models Quinn and co‐workers (1990) weathered oils in a The review of the predictive and theoretical work in controlled environment and correlated the data with section 2 above reveals those air‐boundary‐layer equations developed starting with Fickʹs diffusion law concepts that are limited and cannot accurately and the Clausius‐Clapeyron equation. Crude oil was explain long‐term evaporation. Fingas conducted a divided into a series of pseudo fractions by boiling series of experiments over several years to examine the point. Each fraction was taken to be equivalent to an n‐ concepts (Fingas 1998, 2011). paraffin. The n‐paraffin distributions of a number of naturally weathered crude oils were determined by Wind Experiments capillary gas‐liquid chromatography. The actual A simple experiment to determine whether or not oil measured evaporation was compared with those evaporation is air‐boundary‐layer regulated is to generated by computer simulation of weathering. measure if the evaporation rate increases with wind as Bobra (1992) conducted laboratory studies on the predicted by equations (2) and (6) above. Experiments evaporation of crude oils. The evaporation curves for on the evaporation of oil with and without winds were several crude oils and petroleum products were conducted with ASMB (Alberta Sweet Mixed Blend), measured under several different environmental gasoline, and with water. Water formed a baseline conditions. These data were compared to the equation data set since this is the substance being compared.4 developed by Stiver and Mackay (1984). The equation Regressions on the data were performed and the used was: equation parameters calculated. Curve coefficients are the constants from the best fit equation [Evap = a ln(t)], FV = ln[1 + B(TG/T) θ exp(A ‐ B To/T)] {T/BTG} (32) t=time in minutes, for logarithmic equations or Evap=a where FV is the fraction evaporated, TG is the gradient √t, for the square root equations. Oils such as diesel of the modified distillation curve, A and B are fuel with fewer sub‐components evaporating at one dimensionless constants, To is initial boiling point of time, have a tendency to fit square root curves (Fingas the oil and θ is the evaporative exposure as previously 2011, Ŀi et al. 2004). While data were calculated defined. The constants for the above equation and the separately for percentage of weight lost and absolute results from several comparison runs were carried weight, the latter are usually used because it is more out.21 The agreement between the experimental data convenient. The plots of wind speed versus the and the equation results were poor in most cases. This evaporation rate (as a percentage of weight lost) for comparison showed that the Stiver and Mackay each oil type are shown in Figures 1 to Figure 3. These equation predicts the evaporation of most oils figures show that the evaporation rates for oils and relatively well until time approaches 8 hours, after that even the light product, gasoline, are not increased with it over‐predicted the evaporation. The ʹovershootʹ increasing wind speed. The evaporation rate after the could be as much as 10% evaporative loss at the 24‐ 0‐wind value is nearly identical for all oils, resulting hour mark. This is especially true for very light oils. from the stirring effect on the oil which increases the The Stiver and Mackay equation was also found to diffusion rate to the surface. Stirring will increase the under‐predict or over‐predict the evaporation of oils diffusion and therefore the evaporation rate. The oil in the initial phases. Bobra also noted that most oil evaporation data can be compared to the evaporation evaporation follows a logarithmic curve with time and of water, as illustrated in Figure 4. These data show that a simple approach to this was much more the classical relationship of the water evaporation rate accurate than using equation (30). correlated with the wind speed (evaporation varies as

109 101 www.jpsr.org Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013

U0.78, where U is wind speed). This comparison shows IN THE EVAPORATION RATE OF WATER WITH WIND that the oils studied here are not air boundary‐layer VELOCITY. THIS IS TYPICAL OF AIR‐BOUNDARY‐LAYER REGULATION. COMPARE FIGURE 3 WITH OIL EVAPORATION regulated. IN FIGURES 1 AND 2 WHICH DO NOT SHOW THIS TREND OF 100 VARIANCE WITH WIND VELOCITY.

80 Figure 4 shows the rates of evaporation compared to the wind speed for all the liquids used in the study, as

d 60 well as the evaporation rates of all test liquids versus rate o p

a wind speed. The lines shown are those calculated by v E

t 40 n

e linear regression. This clearly shows that water c

er wind = 0 m/s P wind = 1 m/s evaporation rate increased, as expected, with 20 wind = 1.6 m/s wind = 2.1 m/s wind = 2.6 m/s increasing wind velocity. The oils, ASMB (Alberta

0 Sweet Mixed Blend) and gasoline, do not show rises with increasing wind speed.

25 0 20 40 60 80 100 120 140 160 180 Time - minutes FIG. 1 EVAPORATION OF GASOLINE WITH VARYING WIND Gasoline .) n VELOCITIES. THIS FIGURE ALSO SHOWS THAT THERE IS i 20 m n l

LITTLE VARIATION WITH WIND VELOCITY EXCEPT IN GOING / %

FROM THE 0‐WIND‐LEVEL UP TO THE OTHERS. THIS IS DUE r

.o 15

TO THE STIRRING EFFECT OF WIND AND NOT AIR‐ n i m

BOUNDARY LAYER REGULATION. / % 50 ( 10 Rate Water 40 n ASMB atio r 5 d 30 apo v te a E r o p a

v FCC Heavy Cycle E 20 t

n 0 e c r

e wind = 0 m/s P wind = 1 m/s 10 wind = 1. 6 m/ s wind = 2. 1 m/ s 012 wind = 2. 6 m/ s Wind Velocity - m/s 0 FIG. 4 CORRELATION OF EVAPORATION RATES AND WIND VELOCITY. THE LINES ARE DRAWN THROUGH THE DATA 0 500 1000 1500 2000 2500 POINTS FROM EXPERIMENTAL VALUES. THIS CLEARLY Time - minutes SHOWS NO CORRELATION OF OIL EVAPORATION RATES FIG. 2 VAPORATION OF ALBERTA LIGHT CRUDE OIL WITH WITH WIND VELOCITY AND THE STRONG AND EXPECTED VARYING WIND VELOCITIES. THIS FIGURE SHOWS THAT HIGH CORRELATION OF WATER WITH WIND VELOCITY. THE THERE IS LITTLE VARIATION WITH WIND VELOCITY EXCEPT WATER EVAPORATION LINE IS MOVED TO FIT ON THE IN GOING FROM THE 0‐WIND‐LEVEL UP TO THE OTHERS. VERTICAL SCALE, BUT OTHERWISE IS UNALTERED. THIS IS DUE TO THE STIRRING EFFECT OF WIND AND NOT AIR‐BOUNDARY LAYER REGULATION. These experimental data show that oil is not air 100 boundary‐layer regulated. It should be noted that the Increasing wind speed air movement effect at the lowest level is a stirring 80 effect which increases the diffusion of the components in the oil and thus the evaporation. d e

t 60 a r o

p Saturation Concentration a v E t

n 40 e

c An important concept of evaporation regulation is that r e P of saturation concentration, the maximum 20 wind = 0 m/s wind = 1 m/s concentration of a substance is soluble in air. The wind = 1.6 m/s wind = 2.1 m/s wind = 2.6 m/s saturation concentrations of water and several oil regression line 0 components are listed in Table 1 showing that 0 20 40 60 80 100 120 140 160 180 saturation concentration of water is less than that of Time - minutes common oil components (Fingas, 2011). The saturation FIG. 3 EVAPORATION OF WATER WITH VARYING WIND VELOCITIES. THIS FIGURE SHOWS DRAMATIC DIFFERENCES concentration of water is in fact, about two orders of

110 102 Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013 www.jpsr.org magnitude less than the saturation concentration of temperature in degrees Celsius and t is the time in volatile oil components such as pentane. This further minutes. explains why even light oil components have little Distillation data were correlated to the evaporation boundary layer limitation. rates determined by experimentation. The optimal Development of Generic Equations Using Distillation point was found to be 180oC by using peak functions. Data The percent mass distilled at 180 degrees was used to calculate the relationship between the distillation The evaporation equations for oils show unique values and the equation parameters. The equations differences for oils under the same conditions, used were derived from correlations of the data. The implying that unique equations may be needed for data from those oils that were better fitted with square each oil and this fact is a significant disadvantage to root equations ‐ diesel, Bunker C light and FCC Heavy practical end use. A method to accurately predict Cycle‐were calculated separately. The equations evaporation by means of other readily‐available data derived from the regressions are as follows: is necessary (Fingas 1999). Findings show that distillation data can be used to predict evaporation. For most oils that follow a logarithmic equation: Distillation data are very common and often the only Percentage evaporated= 0.165(%D) ln(t) (34) data used to characterize oils. This is because the data are crucial to operating refineries. Crude oils are For oils that follow a square root equation such as sometimes priced on the basis of their distillation data. diesel fuel:

Table 1 Saturation Concentration of Water and Hydrocarbons Percentage evaporated = 0.0254(%D)√t (35) Substance Saturation Concentration * where %D is the percentage (by weight) distilled at in g/m3 at 25oC 180oC. These equations can be combined with the water 20 n-pentane 1689 equations generated in previous work to account for hexane 564 the temperature variations (Fingas 2011): cyclohexane 357 benzene 319 For oils (most oils and petroleum products) that follow n-heptane 196 a logarithmic equation: methylcyclohexane 192 toluene 110 Percentage evaporated = [.165(%D) + .045(T‐15)]ln(t) ethybenzene 40 p-xylene 38 (36) m-xylene 35 o-xylene 29 For oils like diesel fuel that follow a square root *Values taken from Ullmann's Encyclopedia equation: Oils and diesel‐like fuels evaporate as two distinct Percentage evaporated = [.0254(%D) + .01(T‐15)]√t types, those that evaporate as a logarithm of time and (37) those that evaporate as a square root of time.1 Most oils typically evaporated as a logarithm (natural) with where %D is the percentage (by weight) distilled at o time. Diesel fuel and similar oils, such as jet fuel, 180 C. kerosene and the like, evaporate as a square root of A large number of experiments were performed on time. The reasons for this are simply that diesel fuel oils to directly measure their evaporation curves. and such like have a narrower range of compounds Examples of empirical equations obtained are given in which evaporating at similar rates, yield rates which Table 2. together sum as a square root.

The empirically measured parameters at 15oC were Complexities to the Diffusion-Rgulated correlated with both the slopes and the intercepts of Model the temperature equations. Full details of this Oil Thickness correlation are given in the literature (Fingas 2011). For most oils and petroleum products, the variation Studies show that under diffusion regulation very with temperature resulting equation is: thick slicks (much more than about 2 mm) evaporate slower than other slicks (Fingas 2011). This is due to Percentage evaporated = [B + 0.045(T‐15)]ln(t) (33) the increased path length that volatile components where B is the equation parameter at 15oC, T is must diffuse in a thicker slick. This can certainly be

111 103 www.jpsr.org Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013 confused with air‐boundary‐layer regulation. most significant change that occurs in an oilʹs Experiments by the present author studied the effect composition. Many models in the decade after 1984 of thickness on the evaporation of a light crude oil, use the Stiver and Mackay (1984) approach. Currently, Alberta Sweet Mixed Blend (ASMB) crude oil. The more models employ equations such as found in Table equations noted in Table 2 were all measured at a slick 2. thickness of 1.5 mm which is typical of actual at sea Table 2 Sample of Empirical Equations values (Fingas 2011). The best curve fit is a square root of Oil Evaporation function from which a correction can be given for Oil thickness. Type Equation Alaska North Slope %Ev = (2.86 + .045T)ln(t) Corrected equation factor = equation factor +1 ‐ 0.78 Alberta Sweet Mixed Blend %Ev = (3.24 + .054T)ln(t) *√t (38) Arabian Medium %Ev = (1.89 + .045T)ln(t) Where the corrected equation factor is the factor Arabian Heavy %Ev = (2.71 + .045T)ln(t) corrected for the appropriate slick thickness, the Arabian Light %Ev = (3.41 + .045T)ln(t) logarithmic equation factor is that noted in Table 2, Barrow Island, Australia %Ev = (4.67 + .045T)ln(t) and t is the slick thickness in mm. This equation is true Boscan, Venezuela %Ev = (‐0.15 + .013T)t for thickness values above 1.5 mm at which the Brent, United Kingdom %Ev = (3.39 + .048T)ln(t) original equations were measured. Bunker C ‐ Light (IFO~250) %Ev = (.0035 + .0026T)t Bunker C ‐ long term %Ev = (‐.21 + .045T)ln(t) The Bottle Effect Bunker C (short term) %Ev = (.35 + .013T)t Another confusing phenomenon to understanding California API 11 %Ev = (‐0.13 + .013T)t evaporation is the bottle effect. If all the evaporating Cano Limon, Colombia %Ev = (1.71 + .045T)ln(t) oil mass is not exposed, such as in a bottle, more oil Chavyo, Russia %Ev = (3.52 + .045T)ln(t) vapors than those can readily diffuse through the air Cold Lake Bitumen, AB Canada %Ev = (‐0.16 + .013T)t layer at the bottle mouth may yield a partial or Delta West Block 97, USA %Ev = (6.57 + .045T)ln(t) temporary air‐boundary‐layer regulation effect which Diesel ‐ long term %Ev = (5.8 + .045T)ln(t) may end when the evaporation rate of the oil mass is Diesel Fuel short term %Ev = (0.39 + .013T)t lower than the rate at which the vapors can readily Ekofisk, Norway %Ev = (4.92 + .045T)ln(t) diffuse through the opening. Such effects could occur Federated, AB, Canada %Ev = (3.47 + .045T)ln(t) in reality in situations such as oil under ice, partially Fuel Oil #5 %Ev = (‐0.14 + .013T)t exposed to air or when a thick skin forms over parts of Gasoline %Ev = (13.2 + .21T)ln(t) the oil, blocking evaporation. Gulfaks, Norway %Ev = (2.29 + .034T)ln(t) Hout, Kuwait %Ev = (2.29 + .045T)ln(t) Jumps from the 0‐wind Values IFO‐180 %Ev = (‐0.12 + .013T)t Experimentation shows that studies of oil evaporation Isthmus, Mexico %Ev = (2.48 + .045T)ln(t) at no turbulence or air flow indicate a slight decrease Jet A1 %Ev = (.59 + .013T)t in evaporation rate from those experiments carried out Komineft, Russian %Ev = (2.73 + .045T)ln(t) with slight air movement such as found in an ordinary Lago, Angola %Ev = (1.13 + .045T)ln(t) Lago Treco, Venezuela %Ev = (1.12 + .045T)ln(t) room.22 This is due to the slight stirring of the oil mass which increases the diffusion rate somewhat. Tests of Maya, Mexico %Ev = (1.38 + .045T)ln(t) Nugini, New Guinea %Ev = (1.64 + .045T)ln(t) this phenomenon indicate that further increases in Sahara Blend, Algeria %Ev = (0.001 + .013T)t evaporation rate do not occur with increased air Sakalin, Russia %Ev = (4.16 + .045T)ln(t) movement or turbulence, thus confirming that this is a Scotia Light %Ev = (6.87 + .045T)ln(t) phenomenon only at 0‐wind or turbulence conditions. South Louisiana %Ev = (2.39 + .045T)ln(t) Statfjord, Norway %Ev = (2.67 + .06T)ln(t) Use and Comparison of Evaporation Taching, China %Ev = (‐0.11 + .013T)t Equations in Spill Models Troll, Norway %Ev = (2.26 + .045T)ln(t) Evaporation equations are the prime physical change Udang, Indonesia %Ev = (‐0.14 + .013T)t equations used in spill models. A review of the use of West Texas Intermediate %Ev = (2.77 + .045T)ln(t) evaporation algorithms in oil spill models is given in West Texas Sour %Ev = (2.57 + .045T)ln(t) Fingas, 2011. This is because evaporation is often the The comparison of air‐boundary‐layer models with

112 104 Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013 www.jpsr.org the empirical equations leads to some interesting FROM EQUATION 30) AND AN EMPIRICAL CURVE (FROM conclusions on their applicability. Figure 5 shows a TABLE 2). comparison of the prediction of evaporation of diesel The 0‐wind diesel evaporation calculated using an air‐ fuel using an air‐layer‐boundary model and an layer‐boundary model comes closest to the empirical empirical curve. The 0‐wind diesel evaporation curve, however, is of the wrong curvature. The calculated using an air‐layer‐boundary model comes prediction of diesel evaporation using the wind levels closest to the empirical curve, however, prediction is shows errors as great as 100 percent over about 200 of the wrong curvature. The prediction of diesel hours. evaporation using the wind levels shown, results in prediction errors as great as 100 percent over about 200 hours. Figure 6 shows a comparison of the evaporation of Bunker C using two air‐layer‐boundary models and an empirical curve. The 0‐wind evaporation air‐boundary‐layer prediction comes closest to the empirical curve. As most comparisons shown, the evaporation rate up to about 8 hours is similar to the empirical curve. The prediction of Bunker C evaporation using the wind levels shown results in prediction errors as great as 400 percent over about 200 hours (direct differential in percentage differences). These high values of Bunker C FIG 7 A COMPARISON OF THE EVAPORATION OF PEMBINA evaporation as predicted by air‐boundary‐layer CRUDE USING AN AIR‐LAYER‐BOUNDARY MODEL (EQUATION 30), AN ACTUAL ANALYSIS AFTER 30 YEARS AND models with wind conditions are completely AN EMPIRICAL CURVE. THE EVAPORATION RATE UP TO impossible, as shown by extensive experimentation ABOUT 100 HOURS IS SIMILAR TO THE EMPIRICAL CURVE. and field measurements. THE PREDICTION OF LONG‐TERM EVAPORATION USING EVEN SMALL WIND LEVELS SHOWN RESULTS IN PREDICTION 80 Traditional with varying winds ERRORS AS GREAT AS 60 PERCENT OVER ABOUT 10 YEARS. THESE HIGH VALUES OF EVAPORATION AS PREDICTED BY AIR‐BOUNDARY‐LAYER MODELS WITH WIND CONDITIONS 60 d e

t ARE NOT REALISTIC.

ora Traditional no wind p The 0‐wind evaporation prediction comes closest to 40 Eva t Trad. no wind the empirical curve. The prediction of Bunker C Trad. 10 m/s

ercen Trad. 20 m/s

P evaporation using the wind levels shown results in 20 empirical curve Empirical curve Trad. B. no wind prediction errors as great as 400 percent over about Trad B. 10 m/s Trad. B. 20 m/s 200 hours. These high values of Bunker C evaporation 0 as predicted by air‐boundary‐layer models with wind 0 20 40 60 80 100 120 140 160 conditions are completely impossible. As most Time in Days FIGURE 5 A COMPARISON OF THE EVAPORATION OF DIESEL comparisons shown, the evaporation rate calculated FUEL USING AN AIR‐LAYER‐BOUNDARY MODEL (SUCH AS by most means up to about 10 hours is similar to the FROM EQUATION (30) AND AN EMPIRICAL CURVE SUCH AS empirical curve. FROM TABLE 2. Thus there are three major errors resulting from the 80 Traditional with varying winds use of air‐boundary‐layer models, and the first and most important is that they cannot accurately deal 60 with long term evaporation; then, the wind factor ated r Traditional no wind results in unrealistic values and finally, they have not apo

v 40 E Trad. no wind been adjusted for the different curvature for diesel‐like ent

c Trad. 10 m/s r Trad. 20 m/s evaporation. Some modelers have adjusted their Pe 20 empirical curve Empirical curve Trad. B. no wind models using air‐boundary‐layer models to avoid very Trad B. 10 m/s Trad. B. 20 m/s high values at long evaporation times by setting a 0 maximum evaporation value, which works after a 0 20 40 60 80 100 120 140 160 point in time, but does so artificially. Most models of Time in Days any type will require that one sets a maximum rate to FIG 6 A COMPARISON OF THE EVAPORATION OF BUNKER C USING TWO AIR‐LAYER‐BOUNDARY MODELS (SUCH AS avoid over prediction or values over 100%, for

113 105 www.jpsr.org Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013 example. This can be best illustrated using a long term difference from controlled experiments. It is therefor example. A spill in northern Alberta of Pembina oil important to have models that correspond closely to was sampled 30 years after its spill. Analysis shows the actual physics to begin to more accurately model that this was weathered to the extent of 58% (Wang et phenomena such as evaporation. al. 2004). Figure 7 shows the comparison of the actual value, the empirical projection and the air‐boundary‐ REFERENCES layer predicted value, indicating that the air‐ boundary‐predicted value overshoots the estimate by Blokker, P.C., ʺSpreading and Evaporation of Petroleum over 60%, despite the use of only two low wind values Products on Waterʺ, in Proceedings of the Fourth of 2 and 7 m/s. Use of higher wind values increases the International Harbour Conference, Antwerp, Belgium, pp. evaporation to well over 100%. 911‐919, 1964. Bobra, M., A Study of the Evaporation of Petroleum Oils, Conclusions Manuscript Report Number EE‐135, Environment A review of oil evaporation shows that oil evaporation Canada, Ottawa, ON., 1992. is not air‐boundary‐layer regulated. The results of several experimental series have shown the lack of air Brighton, P.W.M., ʺEvaporation from a Plane Liquid Surface boundary‐layer regulation. The fact that oil into a Turbulent Boundary Layerʺ, Journal of Fluid evaporation is not strictly boundary‐layer regulated Mechanics, Vol. 159, pp 323‐345, 1985. implies that a simplistic evaporation equation will Brighton, P.W.M., ʺFurther Verification of a Theory For Mass suffice to describe the process. The following factors and Heat Transfer From Evaporating Poolsʺ, Journal of do not require consideration: wind velocity, Hazardous Materials, Vol. 23, pp. 215‐234, 1990. turbulence level, area, and scale size. The factors significant to evaporation include time and Brutsaert, W., Evaporation into the Atmosphere, Reidel temperature. Publishing Company, Dordrecht, Holland, 299 p., 1982. A comparison of the various models used for oil spill Butler, J.N., ʺTransfer of Petroleum Residues from Sea to Air: evaporation shows that air‐boundary‐layer models Evaporative Weatheringʺ, in Marine Pollutant Transfer, Ed result in erroneous predictions. There are three issues H.L. Windom and R.A. Duce, Lexington Books, Toronto, including that air‐boundary‐layer models cannot pp. 201‐212, 1976. accurately deal with long term evaporation; second, Drivas, P.J., ʺCalculation of Evaporative Emissions from the wind factor results in unrealistic values and finally, Multicomponent Liquid Spillsʺ, Environmental Science and they have not been adjusted for the different curvature for diesel‐like evaporation. There has been some effort Technology, Vol. 16, pp. 726‐728, 1982. made on the part of modellers to adjust air‐boundary‐ Fingas, M., “Evaporation Modeling”, Chapter 9, in Oil Spill layer models to be more realistic for longer‐term Science and Technology, M. Fingas, Editor, Gulf Publishing, evaporation but these may be artificial and result in NY, NY, pp. 201‐242, 2011. other errors such as under‐estimation for long‐term Fingas, M.F., “A Literature Review of the Physics and prediction. Predictive Modelling of Oil Spill Evaporation”, Journal of A diffusion‐regulated model has been presented along Hazardous Materials, Vol. 42, pp.157‐175, 1995. with many empirically‐developed equations for many oils. The equations are found to be of the form shown Fingas, M.F., “Studies on the Evaporation of Crude Oil and in equation (34) Petroleum Products: I. The Relationship between Evaporation Rate and Time”, Journal of Hazardous It is also noted that in terms of diesel fuel and similar oils the curve is different and follows a square root Materials, Vol. 56, pp. 227‐236, 1997. curve as predicted by equation (35). Fingas, M.F., “Studies on the Evaporation of Crude Oil and The most accurate predictions are carried out using Petroleum Products: II. Boundary Layer Regulation”, the empirical equations as noted in Table 2. If these are Journal of Hazardous Materials, Vol. 57, pp. 41‐58, 1998. not available, the parameters can be estimated using Fingas, M.F., “The Evaporation of Oil Spills: Development distillation data as shown by equations (36) and (37). and Implementation of New Prediction Methodology”, in Marine environments are complex, with many Proceedings of The 1999 International Oil Spill Conference,

114 106 Journal of Petroleum Science Research (JPSR) Volume 2 Issue 3, July 2013 www.jpsr.org

American Petroleum Institute, Washington, D.C., pp. London, A 146, pp. 701‐722, 1934. 281‐287, 1999. Technology, Vol. 16, pp. 726‐728, 1982. Goodwin, S.R., D. Mackay and W.Y. Shiu, ʺCharacterization Tkalin, A.V., ʺEvaporation of Petroleum Hydrocarbons from of the Evaporation Rates of Complex Hydrocarbon Films on a Smooth Sea Surfaceʺ, Oceanology of the Mixtures under Environmental Conditionsʺ, Canadian Academy of Sciences of the USSR, Vol. 26, pp 473‐474, 1986. Journal of Chemical Engineering, Vol. 54, pp. 290‐294, 1976. Wang, Z., M. Fingas, C. Yang, B. Hollebone, and X. Peng, Hamoda, M.F., S.E.M. Hamam and H.I. Shaban, “Biomarker Fingerprinting: Applications and Limitations ʺVolatization of Crude Oil from Saline Waterʺ, Oil and for Source Identification and Correlation of Oils and Chemical Pollution, Vol. 5, pp. 321‐331, 1989. Petroleum Products”, in Proceedings of the Twenty‐ Jones, F.E., Evaporation of Water, Lewis Publishers, Chelsea, Seventh Arctic and Marine Oil Spill Program Technical Michigan, 188 p., 1992. Seminar, Environment Canada, Ottawa, Ontario, pp. Li Y.Y., X.L. Zheng, B. Li, Y.X. Ma and J.H. Cao, 103‐124, 2004. “Volatilization Behaviors of Diesel Oil from the Soils”, Yang, W.C. and H. Wang, ʺModelling of Oil Evaporation in Journal of Environmental Science (China), pp. 1033‐1038, Aqueous Environmentʺ, Water Research, Vol. 11, pp. 879‐ 2004. 887, 1977. Mackay, D. and R.S. Matsugu, ʺEvaporation Rates of Liquid Hydrocarbon Spills on Land and Waterʺ, Canadian Journal Merv F. Fingas A scientist in Edmonton, Alberta, Canada, Chemical Engineering, Vol. 51, Canadian Society for Dr. Fingas has a PhD in environmental physics from McGill University (1996), three masters degrees; chemistry (1984), Chemical Engineering, Ottawa, Ontario, pp 434‐439, 1973. business and mathematics (1978), all from University of Monteith, J.L. and M.H. Unsworth, Principles of Environmental Ottawa. He also has a bachelor of science in Chemistry from Physics, Hodder and Stoughton, 2008. Alberta (1974) and a bachelor of arts from Indiana (1968). Quinn, M.F., K. Marron, B. Patel, R. Abu‐Tanbanja and H. He was Chief of the Emergencies Science Division of Al‐Bahrani, ʺModelling of the Ageing of Crude Oilsʺ, Oil Environment Canada for over 30 years in Ottawa, Ontario and is currently work on research in Edmonton, Western and Chemical Pollution, Vol. 7, pp. 119‐128, 1990. Canada. He works here as an independent researcher and he Reijnhart, R. and R. Rose, ʺEvaporation of Crude Oil at Seaʺ, has published more than 800 papers and publications in the Water Research, Vol. 16, pp. 1319‐1325, 1982. field in addition to 7 books prepared on spill topics and Stiver, W. and D. Mackay, ʺEvaporation Rate of Spills of working on 2 others. Hydrocarbons and Petroleum Mixturesʺ, Environmental Dr. Fingas is a member of the American Chemical Society Science and Technology, Vol. 18, pp. 834‐840, 1984. and the American Association for the Advancement of Science. He has been active in the American Society for Sutton, O.G., “Wind Structure and Evaporation in a Testing and Materials and awarded the honor of ‘fellow’ of Turbulent Atmosphere”, Proceedings of the Royal Society of this society.

115 107 Canadian Science Advisory Secretariat Pacific Region Science Advisory Report 2009/046

RECOVERY POTENTIAL ASSESSMENT FOR BASKING SHARK IN CANADIAN PACIFIC WATERS

Figure 1. Basking shark (Courtesy of T.Campbell). Figure 2. Historic ranges of basking shark in the Northeast Pacific (From McFarlane et al. 2009).

Context :

Basking sharks (Cetorhinus maximus) are a large planktivorous shark that ranges across the north-east Pacific and Atlantic Oceans. They were assessed by COSEWIC in 2007 as Endangered. To support decision-making, information is required on the species’ biology, population trends and targets, habitat requirements, threats to the survival or recovery and allowable harm. This Recovery Potential Assessment (RPA) was developed based on a peer-review by the Pacific Scientific Advice Review Committee (PSARC).

December 2009

108 Pacific Region RPA for Basking Shark in Canadian Pacific Waters

SUMMARY

• The Pacific population of basking sharks is designated by COSEWIC as endangered. • Basking sharks are large planktivorous sharks that range across the north-east Pacific. • They are very slow growing, slow to mature, and have low fecundity. • The historic north-east Pacific population is estimated at 1000-2600 animals. • Since 1996 there have been 12 sightings of basking sharks in Pacific Canadian waters. • It is estimated that 200 years are needed before population numbers will return to their unexploited status but only if human induced mortality is zero. If human induced mortality exceeds 11-18 sharks per year the population could be extinct within approximately 30 years. • A short-term goal for the basking shark is to promote the population’s recovery such that it can be downlisted from its current COSEWIC endangered status. An interim goal is to see positive population growth to the number of observations recorded for the 1945-1970 period. A long-term goal is to promote the sustained viability of a naturally-reproducing population.

INTRODUCTION

The basking shark (Cetorhinus maximus Gunnerus, 1765) is the sole member of the family Cetorhinidae belonging to the order Lamniformes. The basking shark is readily distinguished from other sharks by its large size (maximum reported 12.2 m), elongated gill slits, pointed snout, large mouth with minute hooked teeth, caudal peduncle with strong lateral keels, and crescent shaped caudal fin (Figure 1). Colour is typically blackish to grey-brown, grey or blue- grey above and below on the body and fins. The under-surface is sometimes lighter, and it often has irregular white blotches on the underside of the head and abdomen. Internal gill openings have prominent gill rakers.

Basking sharks have been recorded in surface waters ranging from 8 to 24ºC; with the majority of the observations from 8 to 14ºC waters. Recent tagging has largely disproved the longstanding theory that basking sharks ‘hibernate’ in deep water over the winter (Sims et al. 2003). Very little is known regarding the dispersal and migratory patterns of individual basking sharks. Seasonal migrations are suspected to occur from deep to shallow water or from lower to higher latitudes. This inference is based on observations of seasonal changes in abundance on both the Atlantic and Pacific coasts of North America. In the northeast Pacific, more basking sharks were observed in spring and summer off British Columbia and Washington, and more were counted off California in autumn and winter. It has been inferred from these observations that there is a single northeast Pacific population that migrates seasonally (Compagno 2001).

The presence of basking sharks on the ocean surface in areas of high zooplankton concentrations, combined with the anatomical adaptation of specialized gill rakers, suggests that they are primarily planktivores. Stomach content analyses confirm that zooplankton is the preferred prey, but these analyses are based primarily on basking sharks that were active at the surface when they were captured in commercial fisheries. Other information indicates that a wide range of prey sources, aside from zooplankton, may also be utilized.

Basking sharks are known for their tendency to appear seasonally, in large aggregations, in particular localities, and where they are observed intermittently over several months before disappearing again. In British Columbia, anecdotal and newspaper accounts also indicate that several bays and small inlets were noteworthy for the regular occurrence of high densities of basking sharks. These aggregations may reflect some unknown breeding or foraging behaviour.

2 109 Pacific Region RPA for Basking Shark in Canadian Pacific Waters

This recovery potential assessment (RPA) was based on a science peer by the Pacific Scientific Advice Review Committee (PSARC) and based on DFO guidelines for developing an RPA (DFO 2007a). A more detailed account of the assessment and source material for the review is provided by McFarlane et al. (2009).

ASSESSMENT

PHASE I: assessment of status

Range and historical abundance

Basking sharks in British Columbia and California may belong to a single seasonally migrating population. This hypothesis is based on observations that the seasonal disappearance of basking sharks from California waters between May and July coincided with the appearance of basking sharks in relative abundance in British Columbian waters. Further evidence for this relationship is found in the coincidental disappearance after 1993 of the small congregations of basking sharks that were seen only in Clayoquot Sound in British Columbia and in Monterey Bay in California in the early 1990s (COSEWIC 2007). The likelihood that the basking sharks frequenting coastal inlets of British Columbia and the northeast Pacific Ocean in summer are from the same population of animals occupying central California in the fall and winter underlies the importance of considering all of the basking sharks occurring along the coast of North America as a single population.

It is important to note that basking sharks have only been enumerated while visible at the surface. However, the percentage of time spent at the surface is unknown and is likely influenced by prey distribution, weather conditions, and reproductive behaviours. Basking sharks have been known to spend more time at the surface in shallow, inner continental shelf areas than in deep, well stratified waters. This is thought to be due to associated differences in migratory behaviour of zooplankton prey. Therefore, abundance estimates based on daytime surface sightings may under or overestimate shark abundance by at least 10-fold (Sims et al. 2005). Furthermore, those making the sightings are also influenced by weather and sea conditions, and this may bias the interpretation of seasonality of abundance.

Any decline in abundance of basking sharks has been obscured by the unpredictability in the occurrence and numbers of basking sharks visiting the coastal areas in which they are seen. Early accounts from central California mention basking sharks returning every twenty years or so, and the sudden appearance of large numbers inspired the establishment of fisheries in the early 1920s and again in the late 1940s. Whether the high degree of annual variation in basking shark abundance in the northwest coast of North America is related to climate driven changes in sea surface temperature as is suspected for basking sharks off southwest Britain, is unknown (Cotton et al. 2005).

Estimates of past abundances was made possible by examining a variety of historical records including scientific sources, newspapers, government records pertaining to the 1945-1970 eradication program, commercial harvest, and sport fishing. From 1900 to 1970, basking sharks were regularly found in numerous locations along British Columbia’s coast. Within three generations (66-99 years), basking sharks had all but disappeared from areas where they were once abundant. Throughout this period (1900 – 1970), basking sharks were subjected to a commercial harvest, a directed eradication program, incidental catch, and sport harpooning. It

3 110 Pacific Region RPA for Basking Shark in Canadian Pacific Waters has been estimated that the total number of basking sharks killed in Canadian waters (1945- 1970) by eradication is 413, other patrol/eradication methods (200-300), entanglement (400- 1500), and sport kills (50-400) (COSEWIC 2007). Thus, these eradications totalled from 1000 to 2600 animals. At a mortality rate of 40 animals per year (1000 kills in 25 years) and using estimates of annual productivity, it would take 25 years for an initial population of 750 to be diminished to zero. Note that there is no reliable information on trends in abundance to corroborate this inference.

Current abundance

The current abundance of basking sharks in Canada’s Pacific waters is unknown. At present, basking sharks appear infrequently in Pacific waters, with only six confirmed sightings since 19961 and only ten since 1973 (not including Clayoquot Sound). Four of these sightings are from trawl observer records and the basking sharks were likely killed (COSEWIC 2007). Thus, there is no reliable way to estimate the current population size, but the current population is assumed to be very small.

Darling and Keogh (1994) provide a comprehensive list of reliable sightings in Clayoquot Sound. They list 97 sightings in 1992 (27 individual sharks were identified), 54 basking shark observations from a commercial pilot’s flight log (1973-1992, observations in all but 5 years), and six other observations (1988-1991). All observations were from channels and inlets. However, since 1994 there have been no confirmed sightings from Clayoquot Sound (COSEWIC 2007).

In addition to Barkley Sound, Clayoquot Sound, and Rivers Inlet in British Columbia, the only other areas known for basking shark aggregations in the northeast Pacific Ocean are central and southern California. Areas reported to once support abundant populations in the Pacific northeast are shown in Figure 2.

Life history parameters

The life cycle and reproduction of basking sharks are poorly understood. Pairing for reproduction is thought to occur in early summer based on observed courtship behaviour and scarring. The gestation period has been estimated at 2.6 to 3.5 years. Time between successive litters may be two to three years. Longevity is presumed to be approximately 50 years and age at maturity is estimated at 12 to 16 years in males and 16 to 20 years in females. At birth, basking sharks are between 1.5-1.7 m in length; large enough to escape predation by most marine species. Length at maturity is estimated at 4.6 to 6.1 m for males based on clasper development; females are presumed to mature at a larger size than males as in many other shark species.

Estimates of annual productivity measured as the intrinsic rate of population growth range from 0.013 to 0.023 per year. This suggests that the potential for recovery is lower for basking shark than for other Pacific sharks (Smith et al. 1998). The generation time is estimated to range from 22 to 33 years.

1 Since the time of writing of the initial Research Document (McFarlane et al. 2009) two confirmed sightings have been reported in 2008, and an additional four probable sightings in the 2000-2007 period.

4 111 Pacific Region RPA for Basking Shark in Canadian Pacific Waters

Habitat requirements and habitat use patterns

No specific locations have been identified for reproduction, pupping or rearing, although some other shark species are known to mate in northern areas and pup in southern areas. Feeding has been associated with oceanographic fronts which vary both temporally and spatially. There are areas that were once regularly visited by large numbers of basking sharks (e.g. Barkley Sound, Clayoquot Sound, and Rivers Inlet); however, a recovered stock may not return to these areas. Characteristics that might attract particular life stages such as high seasonal food availability or the occurrence of particular behaviours that might indicate reproductive behaviour have not been identified in these areas.

Residences

SARA s. 2(1) defines Residence as “a dwelling-place, such as a den, nest or other similar area or place, that is occupied or habitually occupied by one or more individuals during all or part of their life cycles, including breeding, rearing, staging, wintering, feeding or hibernating.” Currently the policy for designation and protection of residences under SARA is still being developed. The concept of residence as defined above does not explicitly apply to basking shark, which is an open ocean highly migratory fish.

Recovery goal

A short-term goal is to promote the population’s recovery such that it can be downlisted from its current COSEWIC endangered status. An interim goal is to see positive population growth, perhaps reaching the number of observations (average annual kills) recorded for the 1945-1970 period (i.e. 40 per year). A long-term goal is to promote the sustained viability of a naturally- reproducing population.

PHASE II: scope for management to facilitate recovery

Probability of achieving recovery targets

Given the lack of knowledge of current abundance, migratory behaviour and range, and the impacts of past and present human activities, there is great uncertainty in population abundance projections in Pacific Canada.

Three different recovery targets were assessed: 1) 1000 breeding pairs, 2) 30% of number of individuals pre-exploitation, 3) 30% of the initial biomass pre-exploitation. Pre-exploitation is assumed to be before 1920.

If a breeding population currently exists in the northeast Pacific Ocean and if no further human - induced mortality occurs then 1) it will take hundreds of years for the population to recover to 1000 breeding pairs; 2) 200 years are needed before population numbers will return to their unexploited states; and 3) recovery to 30% of the original biomass could happen within 45 years if complete protection is afforded. The likelihood of a rescue effect from US waters is considered low because abundance in US waters is also depleted.

5 112 Pacific Region RPA for Basking Shark in Canadian Pacific Waters

Sources of harm

Basking sharks are particularly vulnerable to human-induced mortality because of their late age of maturity, low fecundity, long gestation period, long periods between gestations, low productivity, sex segregated populations, use of habitat that supports commercial fisheries, lack of fear of vessels, and current small population size. Like whale sharks, Rhincodon typus, basking sharks return regularly to the same seasonal feeding locations. This tendency for high site fidelity makes the species particularly vulnerable to localised depletion. There is considerable doubt whether even moderate exploitation pressure on basking sharks can be sustained (Compagno 2001), and the ramifications for further erosion of already low genetic diversity worldwide is considerable (Hoelzel et al. 2006).

Human-induced mortality in Pacific Canadian waters is primarily from interactions with fishing gears. Records indicate that basking sharks are readily caught in gillnets but are also caught by trawl (bottom, midwater, and shrimp), and easily become entangled in longlines, prawn traps, cod traps, and even herring seines (Wallace and Gisborne 2006). Between 1942 and 1969, several hundred gillnetters annually fished in the areas of Barkley Sound and Rivers Inlet. It is suspected that several hundred sharks (400-1500) may have been killed from entanglement (COSEWIC 2007).

Harassment, sometimes to the point where it is lethal, has also been mentioned in historic anecdotes and this behaviour may reappear if basking sharks return in numbers to Canada’s Pacific coast.

Tourism is not a critical issue for basking sharks at this time but, as has become apparent in the whale watching industry, if basking sharks do increase off the coast of British Columbia, then protocols to minimize disturbance and prevent harassment of the animals may need to be implemented.

Altercations between boat propellers and basking sharks may be common. The habit of the animals of feeding slowly at the surface in shallow water increases the likelihood of this occurring.

The effects of very loud sounds on shark behaviour are not well documented but some recent evidence suggests that they could potentially disrupt normal behaviours such as feeding, mating, and migration. The apparent attraction of basking sharks to boat propellers, possibly to the sound they generate, may contribute to boat-shark interactions that have led to considerable scarring on the fins and snouts of sharks (Darling and Koegh 1994).

Concerns about potential contamination and bioaccumulation of organic pollutants and heavy metals in marine organisms, particularly long-lived cetaceans, also extend to basking sharks. Basking sharks that feed low in the food chain are unlikely to bioaccumulate contaminants at as high a level as are animals such as killer whales that feed from higher trophic levels. However, indirect impacts of pollution on the planktonic food items of basking sharks could cause local depletion of prey species resulting in a shift in habitat use due to food scarcity.

Likelihood that the current quantity and quality of habitat is sufficient

Habitat quality and quantity for the species is not likely to have changed from historical levels. Given the very low numbers and previous wide distribution of individuals, available habitat does

6 113 Pacific Region RPA for Basking Shark in Canadian Pacific Waters not appear to be limiting survival or recovery. Longer-term trends in climate may influence prey availability but recent theoretical work suggests that basking sharks can achieve a net energy gain under moderate concentrations of prey. Fluctuations in abundance or avoidance of historic seasonal areas of surface feeding may be associated with fluctuations in zooplankton abundance as found for basking sharks off west Ireland, or changes in sea surface temperature driven by global weather patterns as observed off southwest Britain between 1988 and 2001 (Cotton et al. 2005).

PHASE III: Scenarios for mitigation and alternatives to activities

Interactions with fisherman have been few in recent years but the numbers are high considering how few basking sharks have been sighted. From 1996 to 2006, only 10 basking shark sightings could be confirmed in the coastal waters of BC and of these 4 were from observer records of the groundfish trawl fishery. Along the coast of California sightings of 24 individuals have been reported and of these 3 were from observer records of the California drift gill net and set net fisheries (COSEWIC 2007). If the abundance of basking sharks in the coastal waters of the Canadian Pacific increased then the conflicts that were seen from the 1920s to the 1960s might return.

Entanglement and bycatch in fishing gears can be expected, and steps taken to reduce the likelihood of damage and loss to both fishermen and sharks should increase the chances of recovery. In order to achieve zero human-induced mortality due to entanglement or incidental bycatch, revision to fishing gear, fishing areas, and/or fishing plans could be required. A change in fishing plans could include changes in time/area closures, automatic closures based on past distribution of the sharks or closures triggered by some level of potential interaction between sharks and fishing gear. Modification to fishing behaviour could minimize entanglement or incidental bycatch. For example, if a basking shark is sighted at the surface, then setting or hauling fishing gear could be avoided until the shark has left the vicinity. Fishing behaviour modification might be achieved through voluntary participation (i.e. education programs), rather than involuntary participation (i.e. fishing plans). This would alleviate the requirement for such a large scale fishery closure.

Because even moderate exploitation of basking sharks probably cannot be sustained, reliable information on the current level of exploitation is essential for planning effective management and conservation strategies. The imprecise reporting of fishery statistics where several species are lumped together as one category, i.e., “other sharks”, can mask reduction in populations of larger, slower growing species like basking sharks, as well as obscuring changes in community structure. Within the cetacean recovery strategies the observer programs are granted priority. A similar approach could be used for basking sharks where observer programs are expanded to all fisheries with the potential to entangle basking sharks, and species identification and reporting in current observer programs are improved.

Reports of harassment and collisions, and more recent observations of scarring on the dorsal fins of basking shark possibly from boat propellers, indicates a high potential for mortality and injury from contact with boats. A program of public education encouraging responsible boat handling in the vicinity of basking sharks similar to the guidelines and best practices for whale watch operators in BC could be adopted, particularly the minimum approach distance of 100 metres/yards recommended for whales and other marine mammals.

7 114 Pacific Region RPA for Basking Shark in Canadian Pacific Waters

Many of the changes and improvements in fisheries data collection and bycatch management mentioned here have been proposed in the draft National Plan of Action for the Conservation and Management of Sharks of February 9, 2007 (DFO 2007b): • Improve the reporting of discarded bycatch and the associated mortality rates in domestic fisheries through better data collection and species identification by at-sea fisheries observers and through mandatory reporting of all bycatch for the commercial and recreational fishing industry; • Continue awareness-raising efforts among commercial and recreational fishers and other resource users about the risks facing certain shark and shark-like species and promote conservation-based release practices to reduce discard mortality; • Encourage the strengthening of regulations of relevant Regional Fisheries Management Organizations with regard to both the handling and release of shark bycatch species and to improve the identification and reporting of bycatch and associated mortality; and • Review the current practices of all commercial and recreational fisheries and implement, where feasible, new rules or technologies with the potential to reduce both the bycatch of sharks and associated mortality.

CONCLUSIONS AND ADVICE

Under the assumption that the Pacific population of basking sharks has declined by 90% from 1920 to 2007, catch scenarios predicted by the production model have had a drastic effect on these large, old, slow-growing sharks. It is estimated that some 200 years are needed before population numbers will return to their unexploited states if human induced mortality is zero. Even if they suffered a low level of human induced mortality (which based on 2007 population estimates would equal 1-2 sharks killed per year), these sharks will never return to their unexploited population level. If these animals are afforded complete protection, it will still take centuries for the population to recover to 1000 breeding pairs. Recovery to 30% of the original biomass could happen within 45 years. If human induced mortality is allowed to approach 50% of natural mortality (11-18 sharks per year) the basking shark population could be extinct within approximately 30 years.

SOURCES OF INFORMATION

Compagno, L.J.V. 2001. Sharks of the world. An annotated and illustrated catalogue of shark species known to date. Volume 2. Bullhead, mackerel and carpet sharks (Heterodontiformes, Lamniformes and Orectolobiformes). FAO Species Catalogue for Fishery Purposes. No. 1, Vol. 2. Rome, FAO. 269 pp.

Cotton P., A. David, W. Sims, S. Fanshawe and M. Chadwick. 2005. The effects of climate variability on zooplankton and Basking shark (Cetorhinus maximus) relative abundance off southwest Britain. Fisheries Oceanography 14(2):151–155.

COSEWIC. 2007. COSEWIC assessment and status report on the Basking shark Cetorhinus maximus (Pacific population) in Canada. Committee on the Status of Endangered Wildlife in Canada. Ottawa. vii + 34 pp. Available at http://www.sararegistry.gc.ca/status/status_e.cfm.

Darling, J.D., and K. E. Keogh. 1994. Observations of Basking sharks, Cetorhinus maximus, in Clayoquot Sound. BC. Canadian Field-Naturalist 108(2): 199-210.

8 115 Pacific Region RPA for Basking Shark in Canadian Pacific Waters

DFO. 2007a. Revised protocol for conducting recovery potential assessments. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2007/039.

DFO. 2007b. Draft National Plan of Action (NPOA) for the Conservation and Management of Sharks, 09 February 2007. Department of Fisheries and Oceans (DFO), Canada. Available at http://www.dfo-mpo.gc.ca/misc/npoa-pan/npoa-shark_e.htm, Accessed November 05, 2007.

Hoelzel, A.R., M.S. Shivji, J.E. Magnussen and M.P. Francis. 2006. Low worldwide genetic diversity in the Basking shark (Cetorhinus maximus). Biology Letters 2 639–642.

MacFarlane, S., J. King, K. Leask, and L. B. Christensen, 2009. Assessment of information used to develop a Recovery Potential Assessment for basking shark Cetorhinus maximus (Pacific population) in Canada. DFO Can. Sci. Advis. Sec. Res. Doc. 2008/071. vi+98p.

Smith, S.E., D. Au and C. Show. 1998. Intrinsic rebound potentials of 26 species of Pacific sharks. Marine and Freshwater Research 49:663-678.

Sims, D.W., E.J. Southall, A.J. Richardson, P.C. Reid and J.D. Metcalfe. 2003. Seasonal movements and behaviour of Basking sharks from archival tagging: no evidence of winter hibernation. Marine Ecology Progress Series 248:187-196.

Sims, D.W., E.J. Southall, G.A. Tarling and J.D. Metcalfe. 2005. Habitat-specific normal and reverse diel vertical migration in the plankton-feeding Basking shark. Journal of Animal Ecology 74: 755-761.

Wallace, Scott and Brian Gisborne. 2006. Basking Sharks: The Slaughter of BC’s Gentle Giants. Vancouver, New Star Books. 88 pp.

9 116 Pacific Region RPA for Basking Shark in Canadian Pacific Waters

FOR MORE INFORMATION

Contact: The Centre for Science Advice Fisheries and Oceans Canada Pacific Biological Station 3190 Hammond Bay Road Nanaimo BC V9T 6N7

Tel: 250-756-7208 Fax: 250-756-7209 E-Mail: [email protected]

This report is available from the:

Centre for Science Advice (CSA) Pacific Region Fisheries and Oceans Canada Pacific Biological Station 3190 Hammond Bay Road Nanaimo, British Columbia Telephone: (250) 756 7208 Fax: (250) 756-7209 E-Mail: [email protected] Internet address: www.dfo-mpo.gc.ca/csas

ISSN 1919-5079 (Print) ISSN 1919-5087 (Online) © Her Majesty the Queen in Right of Canada, 2009

La version française est disponible à l’adresse ci-dessus.

CORRECT CITATION FOR THIS PUBLICATION

DFO. 2009. Recovery potential assessment for basking sharks in Canadian Pacific waters. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2009/046.

10 117 Canadian Science Advisory Secretariat Pacific Region Science Advisory Report 2009/039

RECOVERY POTENTIAL ASSESSMENT FOR WEST COAST TRANSIENT KILLER WHALES

Figure 1 Photo by Dr. Lance Barrett-Lennard Figure 2. Bathymetric map of the Pacific coast of Canada

Context :

West Coast Transient Killer Whales (Orcinus orca) are one of three different assemblages of killer whales in northeastern Pacific Canadian waters. They were first assessed by COSEWIC in 1999 as Special Concern but were reassessed in 2001 as Threatened. This population of killer whales was listed on Schedule 1 under the Species at Risk Act as Threatened in 2003. A draft Recovery Strategy for transient killer whales was prepared by DFO in 2007, but insufficient information was available to set quantitative recovery goals in that document.

SUMMARY

 Mammal-eating ‘transient’ killer whales off Canada’s Pacific coast are listed as Threatened under the Species-at-Risk Act.

 This Recovery Potential Assessment for West Coast Transient (WCT) killer whales is intended to provide a scientific basis for recovery planning and is based on an archive of 219 individually photo-identified WCT whales collected from 1479 sightings between 1974 and 2006.

August 2009

118 Pacific Region Recovery potential assessment for Transient Killer Whales

 These analyses indicate that the WCT population grew rapidly from the mid-1970s to mid- 1990s as a result of a combination of high birth rate, survival, as well as greater immigration of animals into the nearshore study area. The rapid growth of the WCT population in the mid-1970s to mid-1990s coincided with a dramatic increase in the abundance of the whales’ primary prey, harbour seals, in nearshore waters. Population growth began slowing in the mid-1990s and has continued to slow in recent years.

 The recent slowing of WCT population growth suggests a carrying capacity of 250 to 300 whales. The Potential Biological Removal (PBR) is estimated at only 1.60 animals/year implying that the population would decline if human-induced mortality exceeds this rate.

INTRODUCTION

In April 1999, the northeastern Pacific transient killer whale population was designated Special Concern by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC). The status of this population was reassessed in 2001 based on an existing status report (Baird 2001) and an addendum containing updated information (Trites and Barrett-Lennard 2001), and uplisted to Threatened in November 2001. Reasons for this designation were that it is a “small population that eats marine mammals”, and “individuals have high levels of toxic pollutants”. This population became legally listed on Schedule 1 with the proclamation of the Species-at- Risk Act (SARA) in 2003.

As required by SARA, a Recovery Strategy for Transient Killer Whales in Canada was prepared by Fisheries and Oceans Canada (DFO) and posted for public comments in 2007 (Fisheries and Oceans Canada 2007). Once accepted by the Minister of Fisheries and Oceans, an Action Plan will be developed to achieve recovery goals and objectives developed in the Recovery Strategy.

DFO Science has recently established a Recovery Potential Assessment (RPA) process to provide information and science advice for meeting SARA requirements for listed species, and for deciding whether to add species to the list. A RPA is intended to assess current population status, identify the scope of human induced mortality, and describe the characteristics and availability of critical habitat. At the time the Recovery Strategy for transient killer whales was drafted, an RPA was not available. Since that time an assessment of the recovery potential for WCT has been produced (Ford et al., 2007). The content of this document is largely based on that report.

Species biology

The killer whale Orcinus orca is the largest member of the family Delphinidae and one of the most widely distributed mammals in the world. It occurs in all the world’s oceans and most seas, but is most commonly found in coastal waters in high latitude regions. The global population is estimated to be at least 50,000 (Forney and Wade 2006). The killer whale is the apex marine predator in that it has no natural predators, and is capable of feeding on a great diversity of prey, from the largest whales to small schooling fish. However, individual populations of killer whales may have highly specialized foraging strategies and diets.

Three distinct assemblages of killer whales have been described in coastal waters of the northeastern Pacific Ocean. These assemblages, named transient, resident and offshore, differ in diet and foraging behaviour, acoustic behaviour, morphology, and genetic characteristics. Despite having overlapping ranges, these assemblages do not mix and are thus socially and

2 119 Pacific Region Recovery potential assessment for Transient Killer Whales reproductively isolated from each other. Transient killer whales specialize on marine mammal prey, though they occasionally kill and eat seabirds as well. Despite decades of field observations, WCT killer whales have never been observed to prey on fish, whereas resident or offshore killer whales have never been observed to prey on marine mammals. These foraging specializations appear to be fixed behavioural traits maintained by cultural transmission within populations.

The WCT population is the only one known to be present in Canadian waters and is distributed throughout coastal waters of British Columbia. This population is estimated to comprise 219 whales in 1999, based on long-term photo identification studies, though long gaps between sightings of some individuals added considerable uncertainty to this estimate (Ford and Ellis 1999). An additional 100 or so transient killer whales identified off the central coast of California (Black et al. 1997) were in the past considered to be an extension of this population because of acoustical similarities and occasional mixing with WCT individuals in BC waters (Ford and Ellis 1999). However, a recent reassessment indicated that the available evidence was insufficient to warrant inclusion of these whales in the WCT population (Fisheries and Oceans Canada 2007). This is also the case for Gulf of Alaska transients, which are seen occasionally within the range of WCTs but have only been observed to travel in association with WCTs on one occasion.

ASSESSMENT

Killer whales are long-lived animals that have a low reproductive potential. Although few life history parameters are yet available for transient killer whales, values for resident killer whales presented in Olesiuk et al. (2005) may be generally representative. Survival patterns are typical of mammals, being U-shaped with highest mortality rates in very young (neonate) and very old age classes. Survival rates of juveniles and adults are high (0.97-0.99), particularly among mature females and during periods of population growth. During a period of growth in the northern resident killer whale population, females had a mean life expectancy (at age 0.5 yr) of 46 yrs and a maximum longevity of about 80 yrs. Males had a mean life expectancy of 31 yrs, with maximum longevities of 60-70 yrs.

The population dynamics of transient killer whales were estimated using capture-recapture models based on individual photo-identification data collected during 1479 encounters with WCT whales during 1974-2006 (Ford et al. 2007).

Trends and current status

Survival The survival probability, pooled across all individuals was remarkably high, and relatively stable, across the study period. The average annual survival probability had an estimated median of 0.98 (95% probability interval = 0.95-0.99)

Recruitment Recruitment was determined to be high for the first half of the time series and relatively low for the second half of the time series. Recruitment probability was divided for analysis into the four 8-year increments, starting in 1975. The recruitment probabilities for these time blocks were 0.09, 0.09, 0.04 and 0.03. During the final 6 years, when only new calves were documented, the average recruitment was 0.03 (0.01-0.07).

3 120 Pacific Region Recovery potential assessment for Transient Killer Whales

Sex specific survival / recruitment Estimates of survival were relatively imprecise for calves and sub-adult males due to small sample sizes. Nonetheless, survival was notably high and stable for all age / sex classes. Estimates of average survival show the highest rate for adult females (0.98, 0.97-0.99) and juveniles (0.98, 0.95-0.99), with a slightly lower but similar rate for adult males (0.97, 0.94-0.98). The average survival rate was lower for sub-adult males (0.95, 0.87-0.98) and calves (0.92, 0.82-0.97). However, it is clear that recruitment of new non-calves into the population was negligible during the second half of the time series, with almost all the additions coming as first year calves. There is some evidence of a small level of recruitment for other age classes during the first half of the time series, indicating the recruitment of new non-calves into the population. This could reflect movement of non-calf individuals into the study area. However, this may partly reflect the discovery of new individuals in the population with increasing geographical coverage over time.

Population growth High survival and recruitment has led to a growing population. The average annual population growth rate was 1.06 (0.99-1.23), indicating growth of 6% per year. However, there is also evidence of a recent slowing in population growth, mirroring the decline in recruitment during the second half of this time series. This can be seen in the average estimates for growth over the four 8-year increments, starting in 1975: 1.08 (0.96- 1.50); 1.11 (1.02-1.29); 1.02 (0.98-1.10); and 1.02 (0.98-1.07). Growth at the start of the time series was therefore rapid, tracking relatively high rates of recruitment. In the latter half of the time series there was a decreasing growth rate which corresponds to observed decrease in recruitment.

Population abundance Abundance estimates produced from capture probabilities were consistent with these estimates of population growth rate derived from demographics, showing a growing population but a decreasing rate of population growth. Population size was initially very low (posterior median = 24, 95% probability interval = 11-53) in 1974, grew rapidly for the first half of the time series, but growth appears to have slowed towards the end of the series. The population estimate at the end of the time series was 243 (95% probability interval = 180-339) in 2006.

Population growth was described using a stochastic logistic model, in which the rate of population growth is assumed to decrease linearly with increasing abundance. This density- dependent model defined an equilibrium point of K = 262 whales, which may be interpreted as the current carrying capacity for the WCT population. At the end of the time series the population appeared to be close to reaching this equilibrium point. A more thorough review of population status and trends can be found in Ford et.al. (2007).

Recovery potential

If, as the evidence suggests, WCTs are prey limited and are approaching the carrying capacity of their current range, the potential for future population growth is limited without an increase in prey densities. There is currently no indication of an increasing trend in populations of the WCT’s most important prey species, with the exception of the Steller sea lion. Provided there is no decline in prey availability or unexpected source of increased mortality, the population recovery target in the Transient Killer Whale Recovery Strategy – “a stable abundance over the next five years” – is likely achievable. Recruitment in the WCT population has slowed over the past decade, but continued monitoring will be necessary to determine if this trend continues and whether survival decreases in the future due to density dependence. Killer whales are long- lived, upper trophic-level predators with a very low reproductive potential. They also have a

4 121 Pacific Region Recovery potential assessment for Transient Killer Whales tendency to live in very small, behaviourally specialized and reproductively isolated populations. These factors make killer whale populations such as the WCTs highly vulnerable to even minor increases in levels of mortality.

Sources of harm

Since the live-capture of killer whales for public display ended in the mid 1970s, no directed takes of transient killer whales have been recorded in the region. Indiscriminate shooting of killer whales, once common on the coast, now appears to be very rare (Ford et al. 2000). However, a variety of sources of mortality to WCT killer whales could potentially result from human activities. These threats are described in detail in the Recovery Strategy (Fisheries and Oceans Canada 2007), and are summarized below. It should be noted that some of these threats have not been demonstrated to be a direct cause of mortality in WCTs, or in killer whales generally, and may not represent a lethal risk on their own. However, it is possible for multiple stressors to act synergistically to cause stronger negative and possibly lethal effects.

Contaminants Killer whales in coastal waters of the northeastern Pacific carry significant concentrations of Persistent Bioaccumulating Toxins (PBTs) in their tissue. Of greatest concern are polychlorinated biphenyls (PCBs), which are found at extremely high concentrations in WCT killer whales due to their consumption of marine mammals that are already contaminated with PCBs (Ross et al. 2000, 2004). These compounds are not typically acutely toxic, but can potentially have chronic, slow-acting effects as ‘hormone mimics’ or ‘endocrine disruptors’. Although health effects have not been demonstrated in killer whales, high levels of PCBs in harbour seals have been associated with immuno-suppression and endocrine disruption (Mos et al. 2006). Although PCB levels are declining in the environment, recent models suggest that it will take decades before PCB levels in killer whales decline below the thresholds for adverse effects seen in other species (Hickie et al. 2007).

Also of concern are rapidly increasing levels of polybrominated diphenyl ethers (PBDEs), which have recently become widely used as flame retardants in a variety of products. As with PCBs, the potential direct effect of PBDEs on transient killer whale health is not clear, there is growing evidence of endocrine disruption and immunotoxicity in other species (Fisheries and Oceans Canada 2007).

Biological Pollutants Transient killer whales may be at heightened risk to the impacts of exotic diseases or ‘biological pollution’ due to their consumption of marine mammal prey. Transients may be exposed to pathogens such as viruses and bacteria that are endemic to their mammalian prey or from terrestrial sources, such as domestic pets or livestock. Biological pollutants may have direct effects by causing disease in transient killer whales, which may be predisposed to increased risk or severity of infection due to the immunotoxic nature of PBTs found at high levels in transients. Biological pollutants and pathogens could also affect transients indirectly through mortality and subsequent reduced availability of prey species. For example, a widespread epidemic of Morbillivirus among harbour seals, such as that which caused mass mortalities of seals in northwestern Europe, could have serious consequences for WCT killer whales (Fisheries and Oceans Canada 2007).

Acoustic Disturbance There has been increasing concern in recent years about the potential effects of underwater noise on cetaceans. Acoustic disturbance can be of two types: chronic and acute. Chronic noise

5 122 Pacific Region Recovery potential assessment for Transient Killer Whales is primarily associated with motorized vessel traffic of all types, from commercial shipping to whale watching. Chronic noise can result in masking of communication signals used for social contact or behavioural coordination, or interfere with echolocation signals used for navigation and discrimination. Transient killer whales often forage in silence and may rely on passive listening to locate their prey (Barrett- Lennard et al. 1996). Masking effects of increasing background noise could thus reduce their foraging efficiency. Although there is no direct evidence of the effects of high intensity sounds on transient killer whales, by inference from other cetacean species, detrimental effects might be expected.

Physical Disturbance Vessels moving in close proximity have the potential to affect transient killer whales by disrupting behaviours. Although no studies have focused on transients specifically, resident killer whales have been shown to alter their swimming behaviour when approached by boats (Williams et al. 2002). With the increased intensity of whale watching activity in the vicinity of WCT killer whales in some areas, there is a potential for vessels to disrupt hunting behaviour, thereby reducing overall foraging success. Transient attacks on marine mammals are often prolonged and involve energetic, high-speed swimming, and vessels in close proximity can cause the whales to abandon their attack, or provide the prey item with a refuge to escape from the attacking whales.

Collision with Vessels Killer whales are at some risk of injury or mortality as a result of being struck by boats or ships. Although there are no reported cases of transient killer whales being struck by vessels, four such incidents involving resident killer whales have been documented in recent years, two of which were fatal. It is not clear whether differences in swimming and diving patterns between the two types of killer whales make transients more or less vulnerable than resident killer whales to vessel strikes.

Toxic spills There is evidence that killer whales do not avoid toxic spills, as indicated by the behaviour of both residents and transients during the Exxon Valdez oil spill in 1989 in Prince William Sound, Alaska (Matkin et al. 1999), and of residents during a recent diesel spill in August, 2007, in Robson Bight, B.C. In the case of the Exxon Valdez incident, exposure to oil was associated with unprecedented mortality of both transient and resident whales, which probably died from inhalation of toxic petroleum vapours (Matkin et al. 1999). Expanded oil tanker traffic or hydrocarbon exploration and extraction off the west coast would increase the risk of toxic spills and potential injury or mortality to WCT whales.

Changes in Prey Availability Transient killer whales are marine-mammal specialist predators, and potentially could be affected by major changes in prey availability. In western Alaska, there have been sharp declines in abundance of harbour seals, sea lions, and fur seals, and it has been hypothesized that these declines caused transient killer whales in that region to switch to sea otters, a less desirable prey species (Estes et al. 1998). Because of their reliance on highly-specialized foraging strategies that are maintained by social learning and cultural transmission across generations, killer whales likely have a limited ability to shift to efficient hunting of novel prey species. Because WCT killer whales feed on a variety of different marine mammal species, they may not be vulnerable to minor fluctuations in abundance of a particular prey species. However, a wide-scale decline in multiple prey species, as took place in western Alaska, could have significant consequences for WCTs. Marine mammal prey species in the range of WCTs are currently subjected to relatively low levels of human-related mortality, and no significant

6 123 Pacific Region Recovery potential assessment for Transient Killer Whales changes in such mortality rates are anticipated. However, a major change in the marine ecosystem structure off the west coast, possibly resulting from over-harvesting of fish stocks, could affect transients indirectly through effects on their prey.

Scenarios for Mitigation and Alternatives to Activities

This section of the RPA is intended to provide an inventory of all feasible measures to minimize or mitigate the impacts of human activities on WCT killer whales and their habitat. Much of this is addressed in the draft Recovery Strategy for West Coast Transient Killer Whales (Fisheries and Oceans Canada 2007) and in a recent draft assessment of risk to critical habitat identified for resident killer whales (Lee et al. 2007). The following is extracted from these draft documents and presented here for information. Note that these generally include only mitigation measures available in Canadian waters.

Noise Mitigation Military sonar The Department of National Defence (DND) has established protocols to protect marine mammals from disturbance and/or harm from the use of military active sonar. Maritime Command Order 46-13, for marine mammal mitigation, is to avoid transmission of sonar any time a marine mammal is observed within the defined mitigation avoidance zone, which is established specific to each type of sonar. Ship’s personnel receive training in marine mammal identification and detection. All foreign vessels are subject to Canadian regulations while in Canadian waters. However, concern remains regarding compliance by foreign vessels with Canadian regulations and the effectiveness of these mitigation protocols.

Seismic air guns Currently few industrial or scientific seismic surveys are being conducted in western Canadian waters. Some projects involving seismic surveying trigger screening under the Canadian Environmental Assessment Act (CEAA), while others are reviewed regionally by DFO. In 2008, DFO issued a Statement of Canadian Practice on the Mitigation of Seismic Noise in the Marine Environment (DFO 2008), to address concerns regarding the potential impact of seismic use on marine mammals and other marine life. In the Pacific Region, each proposed seismic survey is reviewed by DFO marine mammal experts and mitigation measures are developed based on the species of concern in the area of the survey for each project.

Construction noise Mitigation protocols to prevent exposure of cetaceans to noise associated with construction activities such as dredging and pile driving in the Pacific Region are similar to those for seismic air guns.

Chronic noise There is currently little mitigation of chronic noise in the marine environment that originates from shipping and other marine vessel traffic. Of particular concern is noise caused by whale- watching vessels that are frequently concentrated in prime transient killer whale foraging habitat (e.g., off Victoria, B.C.). However, whale-watching guidelines developed jointly by DFO and the US National Oceanic and Atmospheric Administration have served to restrict the distance and speeds at which boats can approach killer whales, which helps to reduce the level and extent of noise ensonification in the vicinity of whales.

7 124 Pacific Region Recovery potential assessment for Transient Killer Whales

Toxic Spills The Transportation of Dangerous Goods Act regulates handling and transport of toxic substances within Canada, and numerous international, federal and provincial measures are in place for the prevention and management of toxic spills (e.g. Canadian/U.S. spill response plans for trans-boundary waters, Oil and Gas Operations Act, BC EMA). Despite such regulation and preventative measures, spills are frequent along the coast of British Columbia, but most are very small and localized and do not present a major risk to WCT habitat.

Chemical Pollution Numerous national and international regulations and agreements govern the manufacturing and application of many kinds of PBTs, particularly the so-called legacy PBTs such as PCBs. The Stockholm Convention on persistent organic pollutants (POPs) and other UN Protocols aim to reduce global levels of legacy PBTs. Manufacture and availability of toxic chemicals in Canada are managed via listing under Schedule 1 of the CEPA and the BC Environmental Management Act (EMA) has regulations in place for management of contaminants in industrial and municipal effluents and outflows. The Fisheries Act (S. 36) prevents discharge of toxic substances into fish habitat(s), mitigating toxic threats to killer whale prey. Environment Canada is revising their proposed Risk Management Strategy for Polybrominated Diphenyl Ethers, under the Canadian Environmental Protection Act. This strategy supports the ban of several (but not all) of the forms of PBDEs that are known to bioaccumulate in killer whales. Regulations on manufacture of chemicals and vectors of contamination (e.g. sewage outflows) manage toxins in runoff in British Columbia. The BC Ministry of Environment’s storm-water planning, as well as non- governmental programs are in place for education on toxic runoff. For agriculture, the Fertilizers Act manages chemicals and the BC EMA Agricultural Waste Control regulation and Best Agricultural Waste Management Plans manage industry practices specifically.

Habitat requirements

West Coast Transient killer whales occupy a very extensive range. They travel widely throughout all coastal waters between about 47°N and 58°N latitudes, which corresponds to a straight-line distance of 1600 km along the west coast. Although the extent of their range in offshore waters is unknown, WCT whales have been encountered up to 40 km from shore. Transients spend the majority of their time foraging for marine mammal prey, which likely play a major role in determining movement patterns. WCTs may be found year-round in all parts of their overall range, but they rarely remain in any one location for long. Because their hunting strategy depends on stealth, it is likely more profitable to keep moving once potential prey in an area are alerted to their presence.

Threats to Habitat

In this discussion, direct threats to individual whales, such as physical disturbance or vessel collisions, and threats to prey abundance, are differentiated from threats to habitat. As such, the main threats to WCT habitat are underwater noise and toxic spills. Chronic noise from shipping, for example, has the potential to interfere with foraging success by masking sounds needed to detect and localize prey, especially in confined waterways such as the Inside Passage. Highly industrialized and therefore noisy areas such as ports and harbours may be avoided by transient killer whales, but these are very small areas relative to the whales’ entire range. High intensity acute noise from military sonar or seismic exploration could ensonify large areas of habitat, potentially disturbing whales or displacing them from important foraging areas. Major toxic spills, including catastrophic oil spills, could cause extensive contamination of habitat, in

8 125 Pacific Region Recovery potential assessment for Transient Killer Whales addition to direct physical harm to transient killer whales. This risk can be anticipated to increase if oil tanker or barge traffic increases or oil exploration and extraction takes place in WCT habitat. Other forms of pollution that may degrade habitat quality include wastewater effluent near urban areas, which may contain a variety of chemical pollutants, and pesticides and biological pollutants carried in runoff in agricultural areas.

CONCLUSIONS AND ADVICE

An overall recovery goal and set of specific recovery objectives were developed in the draft Transient Killer Whale Recovery Strategy (Fisheries and Oceans Canada 2007). Without any objective means of estimating historical abundance, carrying capacity or biological limiting factors for WCTs, it was not possible to set a quantitative population abundance target for recovery. Instead, the goal of the Recovery Strategy is:

To attain long-term viability of the West Coast transient killer whale population by providing the conditions necessary to preserve the population’s reproductive potential, genetic variation, and cultural continuity.

Provided there is no decline in prey availability or unexpected source of increased mortality, it can be anticipated that the population recovery target of the Transient Killer Whale Recovery Strategy – a stable abundance over the next five years – is achievable

SOURCES OF INFORMATION Baird, R.W., 2001. Status of Killer Whales, Orcinus orca, in Canada. Canadian Field Naturalist. 155(4): 676-701 Barrett-Lennard, L.G., J.K.B Ford, and K.A. Heise. 1996. The mixed blessing of echolocation: differences in sonar use by fish-eating and mammal-eating killer whales. Animal Behaviour, 51:553-565. Black, N.A., A. Schulman-Janiger, R.L. Ternullo, and M. Guerrero-Ruiz. 1997. Killer whales of California and western Mexico: a catalogue of photo-identified individuals. NOAA-TM- NMFS-SWFSC-247. DFO. 2008. Statement of Canadian practice on the mitigation of seismic noise in the marine environment. http://www.dfo-mpo.gc.ca/oceans-habitat/oceans/im-gi/seismic-sismique/ information_e.asp (Accessed June 1, 2009) Estes, J.A., M.T. Tinker, T.M. Williams, and D.F. Doak. 1998. Killer whale predation on sea otters linking oceanic and nearshore ecosystems. Science 282:473-476. Fisheries and Oceans Canada. 2007. Recovery Strategy for the Transient Killer Whale (Orcinus orca) in Canada [Consultation Draft]. Species at Risk Act Recovery Strategy Series. Fisheries and Oceans Canada, Vancouver 42pp. Ford, J.K.B. and G.M. Ellis. 1999. Transients: Mammal-Hunting Killer Whales. UBC Press, Vancouver, British Columbia. 96 pp. Ford, J.K.B., G.M. Ellis, and K.C. Balcomb. 2000. Killer Whales: the natural history and genealogy of Orcinus orca in British Columbia and Washington, second edition. UBC Press, Vancouver, British Columbia. 104 pp.

9 126 Pacific Region Recovery potential assessment for Transient Killer Whales

Ford, J.K.B., G.M. Ellis, and J.W. Durban. 2007. An Assessment of the Potential for Recovery of West Coast Transient Killer whales Using Coastal Waters of British Columbia. DFO Can. Sci. Advis. Sec. Res. Doc. 2007/088. http://www.dfo-mpo.gc.ca/csas/Csas/ Publications/ResDocs-DocRech/2007/2007_088_e.htm (Accessed June 2009) Forney, K. A. and P. Wade. 2006. Worldwide distribution and abundance of killer whales. Pages 145-162 in J. A. Estes, R. L. Brownell, Jr., D. P. DeMaster, D. F. Doak, and T. M. Williams (editors). Whales, whaling and ocean ecosystems. University of California Press, Berkeley, California. Hickie, B.E., P.S. Ross, R.W. Macdonald, and J.K.B. Ford. 2007. Killer whales (Orcinus orca) face protracted health risks associated with lifetime exposure to PCBs. Environmental Science and Technology 41: 6613-6619. Lee, T. et al. 2007. Preliminary risk assessment of threats to the proposed critical habitat of Northern and Southern Resident Killer Whales (Orcinus orca) in British Columbia. Unpubl. report, Fisheries and Oceans Canada, Pacific Biological Station, Cetacean Research Program. Matkin, C.O., G. M. Ellis, E. L. Saulitis, L. G. Barrett-Lennard, and D. R. Matkin. 1999. Killer Whales of Southern Alaska. North Gulf Oceanic Society, Homer, Alaska. Mos, L., B. Morsey, S. J. Jeffries, M. B. Yunker, S. Raverty, S. De Guise, and P.S. Ross. 2006. Chemical and biological pollution contribute to the immunological profiles of free-ranging harbor seals. Environmental Toxicology and Chemistry. 25 (12): 3110–3117 Olesiuk, P.F., G.M. Ellis, and J.K.B. Ford. 2005. Life history and population dynamics of northern resident killer whales (Orcinus orca) in British Columbia. DFO Can. Sci. Advis. Sec. Res. Doc. 2005/045. http://www.dfo-mpo.gc.ca/csas/Csas/Publications/ResDocs- DocRech/2005/2005_045_e.htm (Accessed December 2006) Ross P. S., G. M. Ellis, M. G. Ikonomou, L. G. Barrett-Lennard and R. F. Addison. 2000. High PCB Concentrations in Free-Ranging Pacific Killer Whales, Orcinus orca: Effects of Age, Sex and Dietary Preference. Marine Pollution Bulletin 40(6): 504-515, Ross, P.S., S.J. Jeffries, M. B. Yunker, R. F. Addison, M. G. Ikonomou, and J. C. Calambokidis. 2004. Harbor seals (Phoca vitulina) in Britsh Columbia, Canada and Washington State, USA, reveal a combination of local and global polychorinated buphenyl, dioxing and furan signals. Environmental Toxicology and Chemistry. 23(1): 157-165 Trites, A. W., and L.G. Barrett-Lennard. 2001. "COSEWIC Status Report Addendum on Killer Whales (Orcinus orca)". Williams, R., D.E. Bain, J.K.B. Ford, and A.W. Trites. 2002. Behavioural responses of male killer whales to a ‘leapfrogging’ vessel. J. Cetacean Research and Management 4:305-310

10 127 Pacific Region Recovery potential assessment for Transient Killer Whales

FOR MORE INFORMATION

Contact: John K. Ford Fisheries and Oceans Canada Pacific Biological Station 3190 Hammond Bay Road Nanaimo, BC V9T 6N7 Tel: 250-729-8375 Fax: 250-756-7053 E-Mail: [email protected]

This report is available from the:

Centre for Science Advice (CSA) Pacific Region Fisheries and Oceans Canada Pacific Biological Station 3190 Hammond Bay Road Nanaimo, BC V9T 6N7

Telephone: (250) 756-7208 Fax: (250) 756-7209 E-Mail: [email protected] Internet address: www.dfo-mpo.gc.ca/csas

ISSN 1919-5079 (Print) ISSN 1919-5087 (Online) © Her Majesty the Queen in Right of Canada, 2009

La version française est disponible à l’adresse ci-dessus.

CORRECT CITATION FOR THIS PUBLICATION

DFO. 2009. Recovery Potential Assessment for West Coast Transient Killer Whales. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2009/039.

11 128 Canadian Science Advisory Secretariat Pacific Region Science Advisory Report 2014/047

RECOVERY POTENTIAL ASSESSMENT OF OFFSHORE KILLER WHALES OFF THE PACIFIC COAST OF CANADA

A group of Offshore Killer Whales. (Photo credit: B. Figure 1. Map of the eastern North Pacific showing Gisborne). locations of encounters with Offshore Killer Whales. Grey line denotes Canada’s Exclusive Economic Zone.

Context : When the Committee on the Status of Endangered Wildlife in Canada (COSEWIC) designates aquatic species as threatened or endangered, Fisheries and Oceans Canada (DFO), as the responsible jurisdiction under the Species at Risk Act (SARA), is required to undertake a number of actions. Many of these actions require scientific information on the current status of the species, population or designable unit (DU), threats to its survival and recovery, and the feasibility of its recovery. Formulation of this scientific advice has typically been developed through a Recovery Potential Assessment (RPA) that is conducted shortly after the COSEWIC assessment. This timing allows for the consideration of peer-reviewed scientific analyses into SARA processes including recovery planning. Offshore Killer Whales were assessed by COSEWIC as Threatened in November 2008, and listed as such under SARA in July 2011. Consequently, a recovery strategy was to be developed by July 2013. A DFO recovery team was struck to develop the recovery strategy, and a Technical Workshop to complement the development of that document was held May 27-28, 2013. No prior RPA or critical habitat assessment or analysis has been conducted. In support of the recovery of Offshore Killer Whales by the Minister, DFO Science has been asked to undertake an RPA, based on the National Frameworks (DFO 2007, 2009). The advice in the RPA may be used to inform both scientific and socio-economic elements of a listing decision, as well as development of a recovery strategy and action plan, and to support decision-making with regards to the issuance of permits, agreements and related conditions, as per section 73, 74, 75, 77 and 78 of SARA. The advice generated via this process will also update and/or consolidate any existing advice regarding this species.

October 2014

129 Pacific Region Recovery Potential Offshore Killer Whales

SUMMARY • Offshore Killer Whales (OKWs) in the eastern North Pacific comprise a single population with a known range that includes continental shelf waters from southern California to the eastern Aleutian Islands, Alaska. • Relative to other Killer Whale populations in Canadian Pacific waters, OKWs are rarely encountered. They were first identified off the British Columbia (BC) coast in 1988, and have only been encountered in the region a total of 103 times up to 2012. • OKWs exhibit a diffuse latitudinal shift in their distribution seasonally, with encounters being most frequent off California during the winter months and off Alaska during summer. In BC waters, OKWs have been encountered or detected acoustically in all months of the year, with some evidence of peaks in March, August and December. • Although the preferred habitat of OKWs in Canadian Pacific waters appears to be outer continental shelf waters, they also make occasional forays into protected inside passage waters. • OKWs appear to feed primarily on sharks, including Pacific Sleeper Shark, Blue Shark, and Spiny Dogfish, although some teleost fishes such as Chinook Salmon and Pacific Halibut are also consumed. • Population modeling using photo-identification data indicates that the OKW population is small, with an average annual abundance estimate of 300 (95% Highest Posterior Density Interval (HPDI) = 257–373). • The population trend appears stable, with average annual survival rates of 0.98 (95% HPDI = 0.92–0.99) balanced by annual recruitment rates of 0.02 (95% HPDI = 0–0.07). • Potential threats to OKW habitat include prey limitation, acute and chronic underwater noise, chemical and biological contamination, oil spills, and disturbance. Potential sources of human-caused mortality include entanglement in fishing gear and vessel strikes. • A Potential Biological Removal (PBR) of 0.55 animals/year suggests that the population could sustain very little anthropogenic mortality without declining. • There is no evidence that the small OKW population is habitat- or prey-limited, either over its total range or within Canadian waters. Although the total available biomass of their elasmobranch prey is not known, the three species known to be consumed by OKWs – Pacific Sleeper Shark, Blue Shark and Spiny Dogfish – are widespread and abundant, and there is no indication of any decline in the abundance of these species. • Additional field studies to better document patterns of habitat use and foraging ecology of OKWs are needed in order to identify potential critical habitat and describe its function, features and attributes. • Continued photo-identification efforts will be necessary to improve estimates of population abundance and life history parameters and to monitor future trends.

2 130 Pacific Region Recovery Potential Offshore Killer Whales

BACKGROUND Rationale for Assessment In 2001, the northeastern Pacific Offshore Killer Whale population was designated Special Concern by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC). The status of this population was reassessed in 2008 and uplisted to Threatened. Reasons for this designation were the very small size of the population and its exposure to threats from “high levels of contaminants, acoustical and physical disturbance, and potential oil spills”. This population became legally listed as Threatened on Schedule 1 of the Species at Risk Act (SARA) in 2011. As required by SARA for species of Special Concern, a Management Plan for Offshore Killer Whales in Canada was prepared by Fisheries and Oceans Canada (DFO) in December 2009 (Fisheries and Oceans Canada 2009). With the uplisting of Offshore Killer Whales under SARA in 2011, DFO is now required to develop a recovery strategy, which is a planning document that identifies what needs to be done to arrest or reverse the decline of a species. In order to provide an up-to-date assessment of the population’s status and potential threats to recovery, DFO Science has been requested to prepare a Recovery Potential Assessment (RPA) that will serve as the scientific basis for the development of the recovery strategy. An RPA provides scientific background, identification of threats and probability of recovery of a population that is deemed to be at risk. Specifically, an RPA addresses the 17 tasks identified in the Revised Protocol for Conducting Recovery Potential Assessments (DFO 2009). This document provides an assessment of the distribution, seasonality, foraging ecology and population status of OKWs based primarily on on-going studies by the Cetacean Research Program (CRP), Pacific Biological Station (Nanaimo, BC). This status assessment is followed by the RPA, which address the 17 tasks identified in DFO (2009), as well as 10 additional tasks related to the identification of important habitats and threats to those habitats. For more detailed information, see Ford et al. (2014). Species Biology and Ecology The Killer Whale is the largest member of the family Delphinidae and one of the most widely distributed mammals. It is currently considered a single wide-ranging species, Orcinus orca, with a cosmopolitan distribution in all the world’s oceans and most seas. The Killer Whale is the apex marine predator, capable of feeding on a great diversity of prey, from the largest whales to small schooling fish. It has no natural predators. Different regional populations of Killer Whales are often distinct ecotypes with highly specialized foraging strategies and diets. These ecotypes are often sympatric, sharing the same waters but maintaining social isolation from each other. Recent molecular studies have shown that ecotypes in the North Pacific and Antarctic represent genetically-distinct lineages that may represent distinct species or subspecies. Three lineages of Killer Whales have been found in coastal waters of the northeastern Pacific Ocean. These lineages, named Transient, Resident and Offshore, differ in morphology, social structure, diet, foraging behaviour and acoustic behaviour. Despite having overlapping ranges, these lineages do not mix and are thus reproductively isolated from each other. Transient Killer Whales (also known as Bigg’s Killer Whales) specialize on marine mammal prey, though they occasionally kill and eat seabirds as well. Resident Killer Whales prey mainly on fish, particularly salmon, and some squid. Offshore Killer Whales also feed on fish and may specialize in preying on sharks. Neither Residents nor Offshores have been observed to prey on marine mammals. These foraging specializations appear to be fixed behavioural traits maintained by cultural transmission within populations.

3 131 Pacific Region Recovery Potential Offshore Killer Whales

Killer whales are long-lived animals that have a low reproductive potential. Although the life history parameters of OKWs are not known, they are likely similar to those of Resident Killer Whales, which are well studied. Survival patterns are typical of mammals, being U-shaped with highest mortality rates in very young (neonate) and very old age classes. Survival rates of juveniles and adults are high (0.97–0.99), particularly among mature females and during periods of population growth. Resident females have a mean life expectancy of about 46 years and a maximum longevity of about 80 years. Males have a mean life expectancy of 31 years, with maximum longevities of 60–70 years. Females give birth to their first viable calf at approximately 14 years, and produce an average of 4.7 calves over a 24-year reproductive period. Gestation is 16–17 months and the minimum calving interval is about 3 years (mean = 4.9 years). Females give birth to their last calf at around 40 years and then become reproductively senescent for the remainder of their lives. Calving is diffusely seasonal, with a peak in fall and winter. Killer whale social structure is matrilineal, with social groupings in some populations having highly stable composition with little or no dispersal of individuals from the matriline. Offshore Killer Whales tend to form large groupings occasionally containing over 100 individuals. Given that the total population abundance is estimated to be only 300 animals (see below), a substantial proportion of the population can occur in one aggregation and could be at risk from anthropogenic threats such as a catastrophic oil spill. Data Sources for Assessment Data for assessment of population status and habitat use by Offshore Killer Whales are based primarily on field encounters during which photographs were collected for identification of individuals from natural markings. Studies of Killer Whales using this technique have been undertaken annually in BC since 1973 by researchers with the Cetacean Research Program and more recently by colleagues in adjacent US waters to the north and south. These studies have focused on Resident and Transient Killer Whales, and Offshore Killer Whales have been only rarely and sporadically encountered in the region. Overall, only 137 encounters with OKWs have taken place over the past 40 years of field effort, which represents about 1% of the 10,580 total encounters with Killer Whales. Of these 137 encounters, 103 took place in Canadian waters, 13 to the south as far as California and 21 in Alaska. Passive acoustic monitoring was also used as a supplement to photo-identification data in assessment of seasonal occurrence of OKWs as the data are less subject to temporal biases in effort over the year. Distinctive stereotyped calls produced by Resident, Transient, and Offshore Killer Whales are readily distinguishable (Ford 1991; Deecke et al. 2005; Cetacean Research Program, unpubl. data). A network of 13 autonomous underwater recording moorings deployed off the coast of British Columbia during 2006–2012 collected long-term acoustic data that included 69 days with detections of distinctive OKW vocalizations. Feeding habits were assessed by the collection of prey fragments from the water at the location of predation events. Prey species identification was made by examination of fish scales or genetic analysis of tissue samples. ASSESSMENT Population Status, Trend and Recovery Targets COSEWIC (2008) considered the Offshore Killer Whale population to comprise a single Designatable Unit (DU). The data used in our current assessment support this conclusion and indicate that the OKW population consists of a single network of socially-connected individuals that ranges over continental shelf and nearshore waters off the west coast of North America from southern California to the eastern Aleutian Islands, Alaska. The extent of potential movements outside of this range is unknown. There are no known adjacent populations that could provide a rescue effect should the OKW population decline.

4 132 Pacific Region Recovery Potential Offshore Killer Whales

Estimates of the current abundance of OKWs are based on photographic identification of individuals from natural markings between 1988 – the first year OKWs were encountered – and 2011. We adopted a statistical modeling approach using a Bayesian formulation of the Jolly- Seber mark-recapture model to estimate abundance and population dynamics of OKWs from photo-identification data, which takes into account mortality and recruitment, as well as the proportion of unnamed animals in the population. The annual probability of individuals being identified and estimates of abundance, survival and recruitment are shown in Figure 2. Identification probabilities were very low in some years, particularly due to relatively few encounters during the first half of the time series, which limited our power to precisely estimate abundance and demographic parameters in the early years of the study. Higher identification probabilities allowed more precise inference in the latter half of the time series (year 2000 onwards). During this time, survival rates were estimated to be high, with an average posterior median of 0.98 (95% Highest Posterior Density Interval [HPDI] = 0.92-0.99), with mortality balanced by per capita recruitment at an average rate of 0.02 (95% HPDI = 0-0.07). As a result, the abundance of distinctly marked whales alive in each year was estimated to be stable around an average estimate of 240 whales (95% HPDI = 223 to 258). Estimates of the proportion of whales that were distinctive in reference encounters had posterior medians ranging from 0.68 to 0.86, resulting in an overall sampling distribution for the average proportion centered on 0.80 (95% HPDI = 0.64 to 0.92). After rescaling to include both distinct and non-distinct whales, the average annual population abundance estimate, A, had a posterior median of 300 (95% HPDI = 257-373). With no knowledge of historical population abundance or current carrying capacity of the habitats of Offshore Killer Whales, establishing quantitative recovery targets for the species in terms of abundance is difficult. Recovery Strategies for both Resident Killer Whale and Transient Killer Whales have recovery goals of ensuring their long-term viability through the maintenance of steady or increasing abundance and other population and distribution objectives (Fisheries and Oceans Canada 2007, 2011). The Management Plan for OKWs has as its goal: To maintain a population level that is viable over the long-term within the known range for the northeastern Pacific Offshore Killer Whale population in Pacific waters of Canada. with the two main objectives over the 10 years after finalization of the plan being to: a) Maintain the population at, or above its current level (averaged over 5 years) b) Maintain the population’s current range of occupancy and distribution on the west coast of B.C. Potential natural limiting factors for OKWs are poorly known. As killer whales have no natural predators, populations are likely to be limited ultimately by food limitation. The abundance of the OKW population is small enough so that inbreeding could potentially be an issue affecting fitness, but it is likely that OKWs have outbreeding mechanisms that mitigate this risk as in other killer whale populations. Severe tooth wear is ubiquitous among adult OKWs, but it is not certain if this is a potential factor affecting survival or longevity. Mass strandings and entrapment are potential sources of natural mortality in OKWs. Distribution, Seasonality and Foraging Ecology The locations of encounters with OKWs are shown in Figure 1. These are distributed from the coast of southern California (~33°30’ N) north to Prince William Sound, Alaska (~60° N), and west to the eastern Aleutian Islands (~160° W). These encounters comprise the known current range of the population. The extent of potential occurrence in oceanic waters beyond the continental shelf is unknown. The first few encounters with Offshore Killer Whales in BC were in

5 133 Pacific Region Recovery Potential Offshore Killer Whales

1988s and were in areas that had not received previous study effort (e.g., off the outer coast of Vancouver Island and in Haida Gwaii). OKWs were first observed in the protected inside waters off eastern and southern Vancouver Island in 1992, 19 years after annual survey effort for Killer Whales began in this area. They have been encountered or detected acoustically in these inside waters on at least 31 occasions since then. It seems likely that the typical range of OKWs is in outer continental shelf waters, but that their range changed in the early 1990s to include periodic visits to protected nearshore areas. It is unknown whether this change reflects a return to part of their historical range, an expansion of their current habitat, or a shift inshore and away from other habitats. Identification Survival 0.8 1.0 t p t φ 0.6 0.8

0.4 0.6

0.2 0.4 Survival Probability, Survival Identification Probability, Identification

0.0 0.2 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010

Recruitment Abundance

1.0 400.0 t r t

N 300.0

0.5

200.0 Abundance, Abundance, Recruitment rate, Recruitment

0.0 100.0

1990 1995 2000 2005 2010 1990 1995 2000 2005 2010

Figure 2. Estimates of the probability of identification (pt), per-capita recruitment rate (Nt) survival rate * (φt),and abundance of distinctive OKW individuals (Nt), for each year 1989-2011 . Vertical lines represent the full range of the posterior distribution for each parameter, circles represent the posterior median and horizontal lines represent the average levels over the time series. *Estimates from years 1988 and 2012 were omitted because identification probability was fixed in the model to ensure parameter identifiability.

6 134 Pacific Region Recovery Potential Offshore Killer Whales

Figure 3. Distribution of encounters with OKWs (red dots; 1988–2012;) and numbers of days in which OKWs were detected acoustically at fixed monitoring sites (open circles; 2006–2012) in Pacific Canadian waters (indicated by the EEZ line). Also shown are depth isobaths to indicate the continental shelf break. Depth contours are indicated with shades of blue: the 200 m isobaths is light blue, 500 m medium blue, and 1000 m dark blue.

The locations of encounters and numbers of acoustic detections of OKWs in Canadian waters are illustrated in Figure 3. Encounters were scattered widely off the coast, although some concentrations are apparent. Clusters of encounters are located off southeastern and northeastern Vancouver Island, which is at least partly due to observer effort; these waters are frequented by numerous whale watch vessels over much of the year. A year-round network of hydrophones maintained by OrcaLab in Johnstone Strait and Blackfish Sound, northeastern Vancouver Island, has also resulted in many detections (Table 1). As mentioned above, these encounters and detections have all taken place since 1992. Other areas with relatively numerous encounters with OKWs include the banks off the southwest coast of Vancouver Island, the nearshore waters off southeast Moresby Island, Haida Gwaii, and around Langara Island. The southeast Moresby Island and Langara Island areas have had relatively high levels of observer effort, but the ratio of encounters to effort seems qualitatively fairly high compared to other areas. Encounters offshore of southwest Vancouver Island are similarly numerous relative to effort. OKWs were found in a variety of marine habitats off the BC coast, from deep oceanic waters beyond the shelf break to the heads of narrow inlets and bays. Judging from encounter rates relative to observer effort (which, as indicated above, is mostly unquantifiable), waters over the outer continental shelf waters and slope may be particularly important habitat for OKWs. Clusters of encounters near Langara Island, in western and eastern Hecate Strait, and off southwest Vancouver Island are all in relatively close proximity to the continental shelf margin or to Moresby Trough, a deep canyon that extends into Hecate Strait from the southwest.

7 135 Pacific Region Recovery Potential Offshore Killer Whales

Offshore Killer Whales appear to exhibit a latitudinal shift in their distribution seasonally, although this shift is rather diffuse. Overall, OKW encounters are most frequent in California during the winter months and in Alaska during summer. In BC, OKWs have been encountered or detected acoustically in all months of the year, with some evidence of peaks in March, August and December. Although the foraging ecology of OKWs is poorly known, it appears likely that they are fish feeders with a potential specialization on sharks. Elasmobranchs were the predominant prey documented in observed predation events by OKWs. Of 40 prey items identified, 37 (93%) were sharks and only 3 (7%) were teleost fishes (Chinook Salmon, Oncorhynchus tshawytscha). Of the elasmobranchs, Pacific Sleeper Shark (Somniosus pacificus) was most common (68%), with Spiny Dogfish (Squalus suckleyi) and Blue Shark (Prionace glauca) together representing less than one-third of observed prey. There is evidence from stomach contents that Pacific Halibut (Hippoglossus stenolepis) is also consumed. A preponderance of sharks in the diet has been hypothesized to be a cause of the severe tooth wear that is pervasive in the population. Ford et al. (2011a) proposed that the hardened dermal denticles (placoid scales) embedded in the skin of sharks cause abrasion of the teeth during prey handling and consumption, leading to the pattern of tooth wear seen in OKWs. Habitat and Residence Requirements Patterns of movement and habitat use of OKWs are expected to be driven by availability of prey. No specific habitats are likely to be used for particular life processes such as mating and calving. Cetaceans are highly mobile and generally do not have “residences” as defined in the Species at Risk Act. The OKW population ranges widely in both outer coast and inside waters off Canada’s west coast. There are no known residence requirements. It is reasonable to assume that the most important property of OKW habitat is the presence of sufficient prey resources to provide for profitable foraging. Although knowledge of the diet of OKWs in Canadian waters is limited, it appears to be dominated by elasmobranch fish, including Pacific Sleeper Shark, Blue Shark and Spiny Dogfish. Chinook Salmon and Pacific Halibut are also known to be consumed by OKWs, but the importance of these prey species is uncertain. The densities of prey species needed to meet the requirements of suitable habitat for OKWs are not known. The known prey species of OKWs are widely distributed in coastal and offshore waters off the Pacific coast of Canada. The Pacific Sleeper Shark is a relatively deep water species found mostly at depths of 150–450 m in continental shelf and slope waters. Fishery bycatches off the BC coast indicate that the species occurs along the shelf break and in deep areas on the shelf such as Dixon Entrance and Moresby Trough in Hecate Strait. This species is also found in some inside passes and channels with particularly deep water (e.g. Johnstone Strait). Blue Sharks are widely distributed throughout the North Pacific in coastal and epipelagic waters beyond the shelf slope. As with Sleeper Sharks, there is no directed fishery for Blue Sharks, but fishery bycatch in BC shows the species occurs in Hecate Strait, Dixon Entrance, along the continental shelf slope, and in oceanic waters. The Spiny Dogfish is found throughout nearshore and continental shelf waters off the BC coast, as well as in oceanic areas beyond the shelf. Potential Threats to Habitat and Limiting Factors Potential threats to habitat from anthropogenic sources are described below. There is no evidence to date that habitat of OKWs has been reduced in quantity or quality by anthropogenic activities such as fisheries. Some habitat areas such as offshore of southwest Vancouver Island likely have higher levels of anthropogenic ambient noise due to increased shipping than was the

8 136 Pacific Region Recovery Potential Offshore Killer Whales case in past decades, but it is not known if such levels are sufficient to cause functional habitat degradation.

Prey Availability If high quality habitat is that which is used regularly for foraging (as is the case for Resident and Transient Killer Whales), reduction in availability of targeted prey species would reduce the value of such habitat. The primary means by which prey reduction could occur is through fisheries. Currently, there is no evidence that the abundance of the three shark species that dominate the known diet of OKWs has declined in recent years or is likely to in the foreseeable future. There is no directed fishery for either Pacific Sleeper Shark or Blue Shark in Canadian Pacific waters, but bycatch of these species does occur in trawl and longline fisheries and is monitored by DFO. Area-weighted catches per unit effort (CPUEs) of Pacific Sleeper Sharks from bycatch monitoring in the Gulf of Alaska are either stable or increasing, depending on the area. There are substantial fisheries bycatches for Blue Sharks in other regions of the North Pacific, but removals are estimated to be 74% of maximum sustainable yield. Rates of removal through bycatch of Blue Sharks in Canadian Pacific waters are considered low, at 20–40 tonnes per year. Recent stock assessments of the Spiny Dogfish in Canadian Pacific waters indicate that relative abundance is stable. In particular, the outside stock (continental shelf waters excluding the Strait of Georgia), which is likely to be consumed by OKWs, is healthy and fishing pressure is considered to be low relative to the estimated size of the population. There is no evidence that the OKW population is currently habitat- or prey-limited, either over its total range or within Canadian waters. With a population of only some 300 animals and a range that encompasses the continental shelf waters for more than 5000 km of coastline, habitat limitation seems highly unlikely. Although the total available biomass of their elasmobranch prey is not known, the three species known to be consumed by OKWs – Pacific Sleeper Shark, Blue Shark and Spiny Dogfish – are widespread and abundant, and there is no indication of any decline in the abundance of these species. However, despite this high biomass there may be unknown factors that limit the availability of these prey species to OKWs and prey limitation is a possible reason for their lack of population growth. It is also possible that there are important prey species of OKWs that have yet to be identified. If OKWs forage preferentially for large elasmobranchs, it is possible that Basking Sharks (Cetorhinus maximus) represented an important food source in the past. Basking Sharks were once abundant in the range of OKWs including Canadian Pacific waters, but decades of exploitation, intentional culling and bycatch mortality in net fisheries almost extirpated them from the region and they remain extremely rare today (COSEWIC 2007). Because killer whale populations are ultimately regulated by food availability, it is important that a better understanding of OKW foraging ecology is attained to assess the potential for prey limitation.

Underwater Noise Given the apparent importance of underwater acoustics for communication and echolocation in Killer Whales, the acoustic environment is considered to be an important feature of critical habitat of Resident and potential critical habitat of Transient Killer Whales. Acoustic properties are no doubt important features of any OKW habitat as well. The acoustic environment of OKW habitat can be affected by two main types of anthropogenic noise, acute and chronic, and these can affect habitats by masking vocalizations or natural ambient sounds that may be used for orientation, communication and echolocation, or by causing behavioural disturbance responses that result in disruption of life processes or avoidance of noisy areas. Acute noise sources include impulsive sounds generated in the mid to low frequency range (< 10 kHz), such as those produced during - seismic surveying, explosions, and construction-related activities such

9 137 Pacific Region Recovery Potential Offshore Killer Whales as pile driving, and non-impulsive sounds with sudden onset and short duration such as mid- frequency military sonar which typically ranges from 2 to 8 kHz. Chronic anthropogenic noise in the ocean is caused primarily by motorized vessels, but other sources such as offshore wind and tidal turbine arrays can also be significant in some regions. Mid-frequency tactical sonar used in navy operations has been observed to cause serious disturbance responses by Resident Killer Whales and the use of acoustic deterrent devices at aquaculture sites has been linked to displacement of Resident Killer Whales from their habitat (Fisheries and Oceans Canada 2009). Potential effects of chronic noise on Killer Whales are not well understood. Increased vessel noise has shown to be associated with the use of higher amplitude vocalizations in Resident Killer Whales, and there is some evidence of reduced foraging efficiency in high-noise habitats. Noise from increased shipping in the world’s oceans has increased ambient noise levels by as much as 12 dB in recent decades. Shipping activity in some areas off the coast of British Columbia is significant and likely to increase. Noise levels are estimated to be particularly high off southwestern Vancouver Island due to cargo vessels transiting between the entrance to Juan de Fuca Strait and Asia or other desitinations. This area is potentially important feeding habitat for OKWs. Underwater noise could also affect OKWs indirectly through effects on their prey. Sharks are sensitive to low frequency sounds and it is possible that changes in shark behaviour or distribution could result from loud anthropogenic noise.

Chemical and Biological Contamination Degradation of water quality due to environmental contaminants poses a potentially significant threat to OKWs, their prey and habitat. The types of contaminants and the pathways by which they may enter Killer Whale habitat and prey, and the potential effects on the health and survival of Killer Whales are discussed in detail in Fisheries and Oceans Canada (2007, 2011). Potential contaminants include persistent organic pollutants (POPs) such as PCBs and PBDEs (polybrominated diphenylethers), dioxins and furans, heavy metals, and DDT. Krahn et al. (2007) provided evidence that levels of PBDEs and DDT were particularly high in OKWs, and suggested that this may be attributable to their presence in coastal California waters, where these chemicals enter the marine environment through agricultural run-off. As high trophic level predators, sharks are particularly susceptible to bioaccumulation and biomagnification of pollutants due to the high lipid content of their liver and their long life span. Levels of POPs and heavy metals such as mercury in shark tissue can exceed recommended levels for human consumption. No assessment has yet been made on heavy metal concentrations in OKWs.

Oil Spills Although oil spills have the potential to cause direct mortality to Killer Whales, a large-scale catastrophic spill would have the potential to render OKWs habitat areas un-inhabitable for an extended period of time. Although the chance of a major spill in outer coast, continental shelf waters is remote, should a spill take place in confined waters such as the narrow inlets and channels occasionally used by OKWs, immediate and acute effects on individuals could occur and the habitat could be seriously degraded. Because OKWs tend to travel in large aggregations, a significant portion of the population could be affected by a single large-scale spill. Currently there are development proposals in environmental review that, if approved, could result in a significant increase in oil tanker traffic in nearshore waters.

Disturbance Disturbance from the close physical proximity of vessels, particularly those involved with whale watching activities, is a major concern for Resident and Transient Killer Whales in nearshore

10 138 Pacific Region Recovery Potential Offshore Killer Whales waters (Fisheries and Oceans 2007, 2011). OKWs are usually encountered in areas outside the current range of most whale watching excursions, but may be an attraction during their visits to inside waters off eastern and southern Vancouver Island. Given the rarity of such visits, targeted vessel disturbance is a negligible concern at present. Sources of Human Induced Mortality and Harm Potential sources of mortality from human causes to Killer Whales generally are described in COSEWIC (2008) and Fisheries and Oceans Canada (2007, 2011), and to OKWs specifically in Fisheries and Oceans Canada (2009). These include vessel strikes, interactions with fisheries (e.g., entanglement in fishing gear), oil spills, and direct killing. Of these potential sources, none has been shown to be the cause of any documented mortalities to OKWs. There is one case of a non-lethal injury to an OKW individual through a likely vessel propeller strike, which severed the dorsal fin (CRP, Pacific Biological Station, Nanaimo, BC, unpublished data). A mass stranding of 20 OKWs took place at Estevan Point, west coast of Vancouver Island, in 1945, but there is no evidence that it was caused by human activities. Mass strandings of Killer Whales are extremely rare. Scenarios for Mitigation and Alternatives to Activities

Prey Depletion There is currently no directed fishery for two of the three shark species – Pacific Sleeper Shark and Blue Shark – that comprise the majority of known OKW prey. However, these two species are taken as bycatch in groundfish longline and trawl fisheries, although the numbers taken are considered low relative to their abundance. The CPUE of bycatch is monitored and management actions could be taken should a decline suggest depletion of these populations. Historically, there have been major fisheries for Spiny Dogfish in Canadian Pacific waters to supply shark livers for Vitamin A production. Currently there is only a relatively small food fishery for this species. There is no immediate conservation concern for stocks of Spiny Dogfish in Canadian Pacific water based on current levels of removals.The stock status of the species is regularly assessed by DFO Science to ensure management of fisheries at sustainable levels. Two other OKW prey species documented in Canadian Pacific waters, Chinook Salmon and Pacific Halibut, are managed through DFO’s Integrated Fisheries Management Plan for Groundfish.

Underwater Noise Military sonar The Department of National Defence (DND) has established protocols to protect marine mammals from disturbance and/or harm from the use of military active sonar and deployment of ordnance. Maritime Command Order 46-13, for marine mammal mitigation, is to avoid transmission of sonar any time a marine mammal is observed within the defined mitigation avoidance zone, which is established specific to each type of sonar. Ship’s personnel receive training in marine mammal identification and detection. All foreign vessels are subject to Canadian regulations while in Canadian waters. There remains some concern regarding compliance by foreign vessels with Canadian regulations and the effectiveness of these mitigation protocols. Seismic testing There are currently few industrial or scientific seismic surveys conducted in western Canadian waters. Some projects involving seismic surveying trigger screening under the Canadian

11 139 Pacific Region Recovery Potential Offshore Killer Whales

Environmental Assessment Act (CEAA), while others are reviewed regionally by DFO. In 2005, DFO developed a draft Statement of Canadian Practice on the Mitigation of Seismic Noise in the Marine Environment, to address concerns regarding the potential impact of seismic use on marine mammals and other marine life. In the Pacific Region, each proposed seismic survey is reviewed by DFO marine mammal experts and mitigation measures are developed based on the species of concern in the area of the survey for each project. Seismic mitigation protocols recommended by DFO Pacific Region are designed to prevent exposure of cetaceans to received sound pressure levels in excess of 160 dB re 1 µPa, which is generally the level at which behavioural disturbance can be anticipated. A slow ramp-up of air gun pressure, or a ‘soft start’, is utilized in the assumption that this will allow cetaceans to leave the area before it is ensonified with intense sound. A safety zone corresponding to the estimated 160 dB re 1 µPa isopleth is established around the sound source, and a marine mammal observer monitors this zone while air guns are operating. If a cetacean enters the safety zone, air gun use is suspended until it has left the zone. While many seismic projects are screened prior to commencement, it is not clear that all projects are assessed for impacts to marine mammals prior to initiation of seismic activity. Also, even with a sound exposure mitigation protocol, OKWs may be difficult to detect by observers in high sea states and thus may be unknowingly exposed to intense sound. Construction noise Mitigation protocols to prevent exposure of cetaceans to noise associated with construction activities such as dredging and pile driving in the Pacific Region are similar to those for seismic air guns. Chronic noise There is currently no mitigation of chronic noise in the marine environment that originates from shipping and other marine vessel traffic.

Biological and Chemical Pollution There are numerous national and international regulations and agreements that govern the manufacturing and application of many kinds of Persistent Bioaccumulating Toxins (PBTs), particularly the so-called legacy PBTs, such as PCBs. The Stockholm Convention on persistent organic pollutants (POPs) and other United Nations Protocols aim to reduce global levels of legacy PBTs. Manufacture and availability of toxic chemicals in Canada are managed via listing under Schedule 1 of the Canadian Environmental Protection Act (CEPA) and the BC Environmental Management Act (EMA) has regulations in place for management of contaminants in industrial and municipal effluents and outflows. The Fisheries Act (S. 36) prevents discharge of toxic substances into fish habitat(s), mitigating toxic threats to killer whale prey. In 2010, Environment Canada published a Final Revised Risk Management Strategy for Polybrominated Diphenyl Ethers, under CEPA. This strategy has provisions for controls of the forms of PBDEs that are known to bioaccumulate in killer whales.

12 140 Pacific Region Recovery Potential Offshore Killer Whales

Regulations on manufacture of chemicals and vectors of contamination (e.g. sewage outflows) manage toxins in runoff in British Columbia. The BC Ministry of Environment’s storm-water planning, as well as non-governmental programs are in place for education on toxic runoff. For agriculture, the Fertilizers Act manages chemicals and the BC EMA Agricultural Waste Control regulation and Best Agricultural Waste Management Plans (BAWMPs) specifically manage industry practices.

Disturbance Disturbance from the proximity of vessels such as those engaged in whale watching is a minor concern for OKWs at present due to the rare and unpredictable occurrence in nearshore waters. The Fisheries Act’s Marine Mammal Regulations legally protects all marine mammals from disturbance and recently drafted amendments will establish legal approach distance thresholds to improve protection. The Species at Risk Act also provides legal protection for listed species including OKWs from disturbance. The ‘Be Whale Wise: Marine Wildlife Guidelines for Boaters, Paddlers and Viewers’ guidance has a range of recommendations to mitigate potential impacts from small vessels.

Fishery Interactions Fishery interactions, such as entanglement in gear or depredation of catches, have not been documented with OKWs. Amendments to the Marine Mammal Regulations under the Fisheries Act will require mandatory reporting of fishery interactions by commercial fishers, including bycatch, entanglement and depredation. Allowable Harm Assessment Due to the small population size of OKWs, any human-induced mortality would be a cause for concern. In order to estimate the level of human-caused mortality that may be allowable without causing serious population-level consequences or preventing recovery, the U.S. National Marine Fisheries Service has devised a means of calculating the Potential Biological Removal (PBR) for marine mammal populations. PBR estimates the maximum number of animals that may be removed per year, excluding natural mortality, while still allowing the population to reach or sustain its ‘optimum sustainable population’ (Wade 1998). PBR is calculated as: 1 PBR ××= FRN min 2 max R where:

= minimum population estimate (20th percentile of estimated population Nmin size; see formula below for its derivation) = maximum theoretical or estimated net productivity of the stock at a Rmax small population size (0.04 as recommended for cetaceans [Wade 1998] and suggested by recent growth rates of Alaskan Resident Killer Whales (Matkin et al. in press) = recovery factor (0.1, based on population abundance, trend and FR vulnerability [Taylor et al. 2003])

To determine Nmin , we used the following formula from Wade (1998):

13 141 Pacific Region Recovery Potential Offshore Killer Whales

Nˆ Nmin = + NCVz 2 )))(1ln(exp( where: Nˆ = point estimate of population size (300 individuals) z = standard normal variate (0.842 for the 20th percentile) NCV )( = coefficient of variation for population estimate (0.1)

From these calculations, Nmin was found to be 276, resulting in a PBR of 0.55 individuals. It is clear that the small OKW population could sustain very little human-induced mortality without declining. Sources of Uncertainty There are numerous sources of uncertainty in this assessment. The OKW population is very seldom encountered compared to other Killer Whale populations in Canadian Pacific waters, and encounters in continental shelf waters off the outer coast, which is likely their primary habitat, are particularly rare. The large group sizes typical of OKWs and their tendency to be widely dispersed when observed in outer coast areas results in a high proportion of incomplete encounters (where an unknown number of animals are not photo-identified). The infrequency of encounters and the dynamic nature of associations within groups present additional difficulties in photo-identifying individuals, particularly with poorly-marked juveniles, and as result a significant proportion of the population is unnamed. Although these uncertainties have been addressed through mark-recapture modeling, there are relatively wide confidence limits in estimates of abundance, survival and recruitment compared to other Killer Whale populations. Life history parameters such as age of maturity, fecundity rates by age, etc., are very poorly known compared with other killer whale populations in the region. Knowledge of the preferred habitats of OKWs while in Canadian waters is poor. There have only been 103 encounters with OKWs over the last 24 years in these waters, and these have been made opportunistically, without dedicated systematic survey effort. As a result, encounter locations are likely highly influenced by spatial and seasonal biases in search effort that are largely unquantified and thus difficult to correct. Although knowledge of the diet of OKWs has improved in recent years, it still remains inadequate to address such questions as the role of different prey species in determining distribution and movement patterns and in potentially limiting population growth. Additional unbiased data on occurrence and feeding habits are needed to better document important habitats and their functions, features and attributes.

CONCLUSIONS AND ADVICE The OKW population ranges widely in continental shelf waters from southern California to the eastern Aleutian Islands, and may occur in Canadian Pacific waters in any month of the year. Recent evidence suggests that this population feeds primarily on sharks, including Pacific Sleeper Shark, Blue Shark, and Spiny Dogfish, although some teleost fishes such as Chinook Salmon and Pacific Halibut are also consumed. Population modeling using photo-identification data indicates that the OKW population is small, with an average annual abundance estimate of 300 (95% Highest Posterior Density Interval (HPDI) = 257–373). The population appears stable, with average annual survival rates of 0.98 (95% HPDI = 0.92–0.99) balanced by annual recruitment rates of 0.02 (95% HPDI = 0–0.07). Potential threats to OKW habitat include prey limitation, acute and chronic underwater noise, chemical and biological contamination, oil spills, and disturbance. Potential sources of human-caused mortality include entanglement in fishing

14 142 Pacific Region Recovery Potential Offshore Killer Whales gear and vessel strikes. A Potential Biological Removal (PBR) of 0.55 animals/year suggests that the population could sustain very little anthropogenic mortality without declining. There is no evidence that the small OKW population is habitat- or prey-limited, either over its total range or within Canadian waters. Although the total available biomass of their elasmobranch prey is not known, the three species known to be consumed by OKWs – Pacific Sleeper Shark, Blue Shark and Spiny Dogfish – appear to be widespread and abundant. However, catch per unit effort (CPUE) of bycatch of Pacific Sleeper Sharks and Blue Sharks is monitored in fisheries. To support a better understanding of Offshore Killer Whale prey, analyses of the CPUE data should be undertaken to determine if there are trends in abundance of this important prey species. Recovery objectives as described in the Management Plan for Offshore Killer Whales (Fisheries and Oceans Canada 2009) are to 1) maintain the population at or above its current level (averaged over 5 years) and 2) maintain the population’s current range of occupancy and distribution on the west coast of B.C. Given that the abundance trend of OKWs appears to be stable, the first objective would appear to have been attained. Although data are very limited, there does not appear to have been any change in the range of occupancy or distribution of OKWs in BC waters over the past twenty years. Additional field studies to better document patterns of habitat use and foraging ecology of OKWs are needed before critical habitat can be identified. Dedicated systematic vessel surveys and an expanded underwater acoustic monitoring network are required to obtain unbiased data on spatial and seasonal occurrence. Additional acoustic recording instruments should be deployed near the continental shelf slope, an area that may be important habitat for OKWs. Satellite tracking of individual OKWs is a potential means of acquiring additional information on movement patterns and preferred habitat and should be considered. Also, application of an effort model to partially correct seasonal and spatial biases in opportunistic sighting data should be undertaken. Continued photo-identification efforts will be necessary to improve estimates of population abundance and life history parameters and to monitor future trends. Additional prey fragment sampling and fecal sampling should be undertaken to determine diet in different seasons and areas in Canadian waters and to assess the potential for food limitation in OKWs. SOURCES OF INFORMATION This Science Advisory Report is from the May 27th to May 28th, 2013 Recovery Potential Assessment Offshore Killer Whale. Additional publications from this meeting will be posted on the Fisheries and Oceans Canada (DFO) Science Advisory Schedule as they become available. COSEWIC 2007. COSEWIC assessment and status report on the basking shark Cetorhinus maximus (Pacific population ) in Canada. Committee on the Status of Endangered Wildlife in Canada. Ottawa. vii + 34 pp. (Accessed 16 June, 2014). COSEWIC. 2008. COSEWIC assessment and update status report on the Killer Whale Orcinus orca, Southern Resident population, Northern Resident population, West Coast Transient population, Offshore population and Northwest Atlantic / Eastern Arctic population, in Canada. Committee on the Status of Endangered Wildlife in Canada. Ottawa. viii + 65 pp. (Accessed 16 June, 2014). Deecke, V.B., J.K.B. Ford and P. Slater. 2005. The vocal behaviour of mammal-eating killer whales: communicating with costly calls. Anim. Behav. 69: 395-405.

15 143 Pacific Region Recovery Potential Offshore Killer Whales

DFO. 2007. Documenting habitat use of species at risk and quantifying habitat quality. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2007/038. DFO. 2009. Revised protocol for conducting recovery potential assessments. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2007/039. Fisheries and Oceans Canada. 2007. Recovery Strategy for the Transient Killer Whale (Orcinus orca) in Canada. Species at Risk Act Recovery Strategy Series. Fisheries and Oceans Canada, Vancouver, vi + 46 pp. Fisheries and Oceans Canada. 2009. Management Plan for the Offshore Killer Whale (Orcinus orca) in Canada [Proposed]. Species at Risk Act Management Plan Series. Fisheries and Oceans Canada, Nanaimo. iv + 51pp. Fisheries and Oceans Canada. 2011. Recovery Strategy for the Northern and Southern Resident Killer Whales (Orcinus orca) in Canada. Species at Risk Act Recovery Strategy Series, Fisheries & Oceans Canada, Ottawa, ix + 80 pp. Ford, J.K.B. 1991. Vocal traditions among resident killer whales (Orcinus orca) in coastal waters of British Columbia. Can. J. Zool. 69: 1454-1483. Ford, J.K.B., E.H. Stredulinsky, G.M. Ellis, J.W. Durban, and J.F. Pilkington. 2014. Offshore Killer Whales in Canadian Pacific waters: Distribution, seasonality, foraging ecology, population status and potential for recovery. DFO Can. Sci. Advis. Sec. Res. Doc. 2014/088. In press. Ford, J.K.B., G.M. Ellis, C.O. Matkin, M.H. Wetklo, L.G. Barrett-Lennard, and R.E. Withler. 2011. Shark predation and tooth wear in a population of northeastern Pacific killer whales. Aquat. Biol. 11:213-224. Taylor, B.L., M. Scott, J.E. Heyning, and J. Barlow. 2003. Suggested Guidelines for Recovery Factors for Endangered Marine Mammals under the Marine Mammal Protection Act. U.S. Dept. Commer., NOAA Tech. Memo. NMFS SWFSC-354, 6 p. Wade, P. R. 1998. Calculating limits to the allowable human-caused mortality of cetaceans and pinnipeds. Marine Mammal Science, 14: 1-37.

16 144 Pacific Region Recovery Potential Offshore Killer Whales

THIS REPORT IS AVAILABLE FROM THE: Centre for Science Advice (CSA) Pacific Region Fisheries and Oceans Canada 3190 Hammond Bay Road Telephone: 250 756-7208 E-mail: [email protected] Internet address: www.dfo-mpo.gc.ca/csas-sccs/ ISSN 1919-5087 © Her Majesty the Queen in Right of Canada, 2014

Correct Citation for this Publication: DFO. 2014. Recovery Potential Assessment of Offshore Killer Whales off the Pacific Coast of Canada. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2014/047. Aussi disponible en français : MPO. 2014. Évaluation du potentiel de rétablissement des épaulards hauturiers au large de la côte du pacifique du Canada. Secr. can. de consult. sci. du MPO Avis sci. 2014/047.

17 145

Canadian Science Advisory Secretariat Pacific Region Science Response 2015/007 SUFFICIENCY REVIEW OF THE INFORMATION ON EFFECTS OF UNDERWATER NOISE AND THE POTENTIAL FOR SHIP STRIKES FROM MARINE SHIPPING ON MARINE MAMMALS IN THE FACILITIES APPLICATION FOR THE TRANS MOUNTAIN EXPANSION PROJECT

Context Trans Mountain Pipeline ULC (Trans Mountain) is proposing an expansion of its current 1,150 km pipeline system between Strathcona County, AB and Burnaby, BC (the Trans Mountain Expansion Project, hereafter ‘the Project’). On December 16, 2013, Trans Mountain filed a Facilities Application for the Project with the National Energy Board (NEB), pursuant to the National Energy Board Act, and on April 2, 2014, the NEB determined that the Application was complete, thereby commencing a 15-month environmental assessment review under the Canadian Environmental Assessment Act, 2012. As part of the application, NEB included a requirement that the Proponent (Trans Mountain) consider the environmental and socio-economic effects of marine shipping activities that would result from the proposed Project, including the potential effects of accidents or malfunctions that may occur; this was delivered by way of issue # 5 in the NEB’s List of Issues to be considered at the Hearings (NEB 2013a) and was provided to Trans Mountain in a letter outlining its filing requirements with respect to this topic (NEB 2013b). Through scoping, the Proponent has identified underwater noise from Project-related marine vessel traffic as a potential source of sensory disturbance to marine mammals. They consider ship strikes to marine mammals as a potential accident or malfunction that could occur between marine mammals and Project-related traffic. The assessment of potential effects of the increase in Project-related marine vessel traffic is centered on the established in-bound and out-bound marine shipping lanes in the Marine Regional Study Area (Marine RSA) (Figure 1). An increase in marine vessel traffic associated with the Project has the potential to result in sensory disturbance to marine mammals from underwater noise, ranging from auditory injury to behavioural disturbance, and an increased risk of injury and mortality associated with mammal- vessel strikes. Disturbance responses associated with increased Project-related vessel traffic could range from temporary displacement, to reduced foraging efficiency, to disruption of mating and social behaviours. As an Intervenor in the environmental assessment hearing process for the Trans Mountain Expansion Project, Fisheries and Oceans Canada (DFO) will be asked to present evidence at Public Hearings in October 2015 in relation to its expertise on the effects of the Project on fish and fish habitat and aquatic species at risk, the efficacy and adequacy of mitigation and offsetting measures, monitoring and follow-up programs proposed by the Proponent, and the conclusions reached in Facilities Application for the Project. DFO’s Pacific Region Fisheries Protection Program (FPP) is responsible for the review of the marine terminal, and shipping components of the proposed Project. FPP is requesting DFO

February 2015 146 Pacific Region Trans Mountain Expansion Project

Science Branch provide an evaluation of the adequacy of the Proponent’s Facilities Application and supplemental information with respect to risks and potential consequences of underwater noise and marine vessel strikes that may result from Project-related marine shipping on indicator marine mammals in the Marine RSA. FPP is requesting Science advice to assist in the development of DFO’s evidence submission to the National Energy Board for the Trans Mountain Project. DFO Science Branch has been asked to focus its review on direct effects on marine mammals, not including potential effects of underwater noise on prey species, such as fish. The objective of this CSAS Science Response (SR) is to review information provided by the Proponent in the December 2013 Facilities Application, and in supplemental filings with the NEB, and to provide a Science Response to answer the following questions: 1. Is the information provided by the Proponent in the Project Application sufficient for DFO Science to assess underwater noise and marine vessel strikes, and their potential impacts on the marine mammal indicator species identified in the Project Application? 2. Are the methods used to assess the potential effects of underwater noise and marine vessel strikes on the marine mammal indicator species appropriate and executed properly? This Science Response Report results from the Science Response Process of December 5, 2014 on the Sufficiency review of the information on effects of Marine Shipping on Marine Mammals in the Facilities Application for the Trans Mountain Expansion Project.

2 147 Pacific Region Trans Mountain Expansion Project

Figure 1. The Marine Regional Study Area (RSA). (from Trans Mountain Pipeline ULC, 2013. Trans Mountain Expansion Project – An Application Pursuant to Section 52 of the National Energy Board Act, Volume 8A - Marine Transportation).

3 148 Pacific Region Trans Mountain Expansion Project

Background The existing Trans Mountain pipeline (TMPL) system commenced operation in 1953, and transports a range of crude oil and petroleum products from Western Canada to locations in central and southwestern British Columbia (BC), Washington State, and offshore. The proposed Project would create a twinned pipeline, increasing the capacity of the system from approximately 300,000 barrels per day to 890,000 barrels per day. Key Project components include 994 km of new pipeline, reactivation of 193 km of existing pipeline, 12 new pump stations and expansion of existing pump stations and storage tanks, and the addition of three new vessel berths at the Westridge Marine Terminal in Burnaby, BC. For this particular Science Branch review, the marine vessel traffic that will transport the petroleum products is of relevance. The proposed expansion is forecasted to increase marine vessel traffic from 5 tankers per month calling at the Westridge Marine Terminal to approximately 34 tankers per month (i.e., an additional 720 tanker transits each year). At present the maximum size of petroleum tankers that call at the Westridge Terminal are Aframax class which have an average cargo carrying capacity of 750,000 barrels. The maximum size of tankers is not expected to change as part of the Project. These vessels will transit the Marine Regional Study Area (Marine RSA) using existing in-bound and out-bound shipping lanes (Figure 1). It will take each Project-related Marine vessel approximately 12 hours to complete one transit of the Marine RSA, and on average, there will be two transits every 24 hours. This will be in addition to existing traffic in the shipping lanes and other traffic in the Marine RSA (Figure 1). There are 22 species of marine mammals identified by the Proponent as being present in the Marine RSA. The most commonly observed species of toothed whales in the Marine RSA include Killer Whales (Orcinus orca), Harbour Porpoises (Phocoena phocoena), Dall’s Porpoises (Phocoenoides dalli), and Pacific White-sided Dolphins (Lagenorhynchus obliquidens). Critical habitat for the Southern Resident Killer Whale population listed as endangered under the Species at Risk Act (SARA) overlaps almost entirely with the Marine RSA (Figure 2). The Humpback Whale (Megaptera novaeangliae), which is listed as threatened under the SARA, is the most commonly observed baleen whale, and the western-most portion of the Marine RSA overlaps proposed Humpback Whale critical habitat (Figure 2). Other baleen whales including the Minke Whale (Balaenoptera acutorostrata) and Grey Whale (Eschrichtius robustus), as well as the occasional Fin Whale (Balaenoptera physalus) are also observed. The Harbour Seal (Phoca vitulina), and Steller (Eumetopias jubatus) and California (Zalophus californianus) Sea Lions are the most common pinnipeds observed in the Marine RSA, as well as the occasional Northern Elephant Seal (Mirounga angustirostris). Sea Otters (Enhydra lutris) are also occasionally sighted. The Proponent’s assessment of potential effects is on three marine mammal species identified as indicators of toothed whales, baleen whales and pinnipeds in the Marine RSA; respectively these are Southern Resident Killer Whales, Humpback Whales, and Steller Sea Lions. The increase in marine vessel traffic associated with the proposed Project has the potential to result in sensory disturbance to marine mammals from underwater noise, and an increased risk of injury and mortality associated with mammal-vessel strikes. This Science Response examines whether the Proponent has provided sufficient information and conducted appropriate analyses from which to draw conclusions on the effects of increased marine vessel traffic on selected marine mammal indicator species.

4 149 Pacific Region Trans Mountain Expansion Project

Figure 2. Critical Habitat for Southern Resident Killer Whales, proposed Critical Habitat for Humpback Whales and other important areas for marine mammals in the Marine RSA. (from Trans Mountain Pipeline ULC, 2013. Trans Mountain Expansion Project – An Application Pursuant to Section 52 of the National Energy Board Act, Volume 8A - Marine Transportation).

5 150 Pacific Region Trans Mountain Expansion Project

Analysis and Response Science Branch responses to the two sufficiency questions posed by DFO FPP above (see Context) are detailed below. To prepare this response, the following documents from the Proponent were reviewed: 1. Trans Mountain Pipeline ULC. 2013. Trans Mountain Expansion Project – An Application Pursuant to Section 52 of the National Energy Board Act, Volume 8A - Marine Transportation. Submitted to the Secretary of The National Energy Board. 2. JASCO Applied Sciences. 2014. Supplemental Underwater Noise Modelling for Trans Mountain Expansion Project. Prepared for Stantec Ltd. Document 00542, Version 5.0: 51 p. 3. Trans Mountain Pipeline ULC. 2014. Trans Mountain Response to The National Energy Board Information Request No. 1. NEB Hearing Order OH-001-2014. Trans Mountain Document Number: A3W9H8. Submitted to the Secretary of The National Energy Board. 4. Trans Mountain Pipeline ULC. 2014. Trans Mountain Response to Information Request from Val Veirs. NEB Hearing Order OH-001-2014. Trans Mountain Document Number: A3X6V7. Submitted to the Secretary of The National Energy Board. 5. Moffat & Nichol. 2013. TERMPOL 3.2 – Origin, Destination & Marine Traffic Volume Survey. Trans Mountain Expansion Project. Prepared for Trans Mountain. Trans Mountain Document Number: A3S4R7 and A3S4R8. Submitted to the Secretary of The National Energy Board. 6. Trans Mountain Pipeline ULC. 2014. Trans Mountain Response to The National Energy Board Information Request No. 2. NEB Hearing Order OH-001-2014. Trans Mountain Document Number: A3Z4T9. Submitted to the Secretary of The National Energy Board.

Information Sufficiency: Is the information provided by the Proponent in the Project Application sufficient for DFO Science to assess underwater noise and marine vessel strikes and their potential impacts on the marine mammal indicator species identified in the Project Application? The Proponent has provided an extensive review of literature on underwater noise and the potential effects on marine mammals and has referenced publications that evaluate ship strike risk to whales in shipping lanes. However the Proponent’s assessment of effects lacks a standardized structured quantitative or qualitative framework with which to adequately evaluate the impacts of underwater noise and of ship strikes and their potential significance. There are methods available in the literature (e.g., Lawson and Lesage 2012) that provide a framework to systematically identify, describe, and evaluate activities and their potential effects on ecosystems or their components. The inclusion of an appropriate impact assessment framework would facilitate the systematic evaluation of the results and conclusions. Marine Traffic and Ship Strikes Project Application: Volume 8A, Marine Transportation, section 4.3 Effects Assessment – Marine Vessel Traffic operations

Marine mammals are identified as an Environmental and Socio-economic element occurring in the Marine RSA and the Proponent describes effects of the Project on marine mammals in section 4.3.13 as “Accidents and Malfunctions”; and in section 4.3.7.4 (Potential Effects and Mitigation Measures) as a result of normal operations. The Proponent considers ship strikes to be an accidental event based on shipping statistics for the Marine RSA drawn from a TERMPOL

6 151 Pacific Region Trans Mountain Expansion Project

report (Moffat and Nichol 2013; referenced in list of documents above) and DFO’s Marine Mammal Incident Database regarding numbers of reported of ship strikes on whales. Specifically, the Proponent characterizes ship strikes in section 4.3.13 (Accidents and Malfunctions) as: “unplanned events that could result in significant adverse effects to …the environment; however, are unlikely to occur… accidents and malfunctions are predicted to be unlikely for the increased Project-related marine vessel traffic….” And, in section 4.3.13.5.4 Physical Injury or Mortality of a Marine Mammal Due to a Vessel Strike states: “While ship strikes leading to marine mammal fatalities can and do occur, such occurrences are infrequent relative to the number of vessels (of all sizes and classes) on the water”; and “The overall probability of a Project-related vessel striking and injuring a marine mammal is considered low”; and “the frequency of [injury or mortality of a Marine Mammal to a Vessel Strike] is considered accidental and rare for any particular vessel”. The information and analysis provided is insufficient for either a quantitative or a qualitative evaluation of the current rate of ship strikes, or how the rate may increase with the proposed increase in vessel traffic. Neither the analysis of shipping statistics for the Marine RSA, nor analysis of the data from the DFO’s Marine Mammal Incident Database, is used in a manner suitable for assessing the potential effect (exposure or consequence) of ship strike events on marine mammals in the Marine RSA. Specifically: • Ship strikes are known to occur under normal shipping operations. However in many cases ship operators are unaware that a ship strike has taken place, and struck whales are often either not detected and/or sink. As a result, statistics based on recovery of dead whales under-represent the true frequency of ship strikes (Laist et al. 2001; Douglas et al. 2008). Therefore, ship strike data in DFO’s Marine Mammal Incident Database do not represent the true frequency of ship strike occurrences. This uncertainty has not been taken into account. • No information is provided about speed or maneuverability of Project-related ships, distribution of whales in relation to the shipping lanes, or shipping intensity in the Marine RSA. To evaluate the effect of potential ship strikes, related to the Project, on marine mammals, a risk assessment framework that considers the likelihood of a vessel strike to an individual of a specific cetacean population, the size and status of that particular cetacean population, and the propensity of that population for being struck by ships is necessary, but has not been completed. For example, for cetacean species listed as Threatened or Endangered under Canada’s Species at Risk Act, such an assessment would consider the risk of injury or mortality of an individual as a potential threat to the viability of already small populations (Williams and O’Hara 2009). Without such a structured qualitative/quantitative framework, the assessment is insufficient to evaluate the conclusion that the probability of a Project-related vessel striking and injuring a marine mammal is low, or that the occurrence of injury or mortality of a marine mammal would be infrequent. It should also be noted that because there is evidence that ship strikes are known to occur under normal shipping operations, ship strikes could be considered a collateral effect of routine marine shipping, and be treated similarly to disturbance or injury from underwater noise.

7 152 Pacific Region Trans Mountain Expansion Project

Marine Traffic and Underwater Noise Volume 8A, Marine Transportation, section 4.3 Effects Assessment – Marine Vessel Traffic operations subsection 4.3.7.4.2. The Project Application includes a qualitative assessment of “significance” of the potential effect of underwater noise on each of the marine mammal indicator species, as summarized in Tables 4.3.7.9 1(a) and 2(a). The Proponent does not present a quantitative or qualitative framework by which to measure or classify the effect or consequences of the noise. However, the Proponent refers to a study that identifies sound levels that were correlated with specific behavioural changes in Northern Resident Killer Whales (MacGillivray et al. 2012), but does not formally include these data to evaluate Project-related underwater noise and potential impacts on Southern Resident Killer Whales. From MacGillivray et al. 2012: “The study determined that at received sound levels of approximately 64 dB re: HT [Hearing Threshold], killer whales overtly avoided a whale-watching boat, while at received SPLs of approximately 57 dB re: HT, they exhibited subtle avoidance responses.” While this study is identified, the Proponent does not formally incorporate the data from the study or acknowledge that there may be a link between the observed behavioural responses of Northern Resident Killer Whales at these sound levels, and the probability of behavioural responses of Southern Resident Killer Whales at the noise levels that are anticipated as a result of increased Project-related vessel traffic. DFO recognizes that quantifying the effects of individual behavioural response at the population level is difficult. However; there have been important advances in this area that highlight the evidence of physiological responses to increased noise at below threshold levels (Rolland et al. 2012). As well there are recent efforts to develop frameworks to relate noise disturbance to individuals to population level effects (Clark et al. 2009; NRC 2005). Noise impacts on reproductive success and survival are pathways to a population level effect, but these have not been formally considered by the Proponent. Volume 8A, Marine Transportation, section 4.3 Effects Assessment – Marine Vessel Traffic operations, Table 4.3.7.9. The Proponents provide a definition of significance that does not incorporate information or data to allow for a quantitative assessment to determine under what conditions “significant” noise events occur. The Project Application states that significance means: “A high probability of occurrence or a permanent or long-term residual effect of high magnitude that cannot be technically or economically mitigated”. The effects assessment incorporates information from SARA Recovery Strategies and Management Plans and COSEWIC reports for the indicator species (Southern Resident Killer Whales, Humpback Whales, Steller Sea Lions). The effect assessment considers the following information: • whether or not underwater noise is listed as a threat or concern for the species • the length of time an individual might be exposed to noise from a single Project-related ship in the RSA • whether all or only a portion of the species population occurs in the RSA • the proportion of the year that the population resides in the RSA • whether or not Critical Habitat for the species overlaps the RSA • whether or not the acoustic environment is identified as a feature of Critical Habitat for the species

8 153 Pacific Region Trans Mountain Expansion Project

While the use of this information may be appropriate, it is not presented in a qualitative framework that illustrates how these factors were ranked. As presented in Table 4.3.7.9, it is not possible to evaluate the Proponent’s conclusion that the residual effects would be significant for Southern Resident Killer Whales, but not significant for either Humpback Whales or Steller Sea Lions. Volume 8A, Marine Transportation, section 4.3 Effects Assessment – Marine Vessel Traffic operations, subsection, 4.3.7.4.5, Assessment of Potential for Residual Effects of Auditory Injury Potential harm from Project-related marine vessels has been assessed quantitatively by comparing the modelled sound level contours against published acoustic thresholds associated with temporary and permanent hearing threshold shifts (TTS and PTS). The assessment of effects from noise levels at or below the threshold for behavioural disturbance is not sufficient. The assessment considers noise from a single Project-related ship, without taking into account the additive and cumulative effects of existing noise, or increased noise due to Project-related increases in vessel traffic. These omissions inhibit the assessment for potential residual effects from Project-related ship noise on indicator marine mammal species. The limitations in the application of the model will be addressed in the next section.

Model Adequacy and Applicability: Are the methods used to assess the potential effects of underwater noise and marine vessel strikes on the marine mammal indicator species appropriate and executed properly? Ship Strikes While the potential for marine mammal-vessel ship strikes is discussed in the Project Application, the potential risks associated with ship strikes and marine mammals have not been quantified or formally assessed using any existing methodologies (e.g., Vanderlaan et al. 2008; Williams and O’Hara 2010). In general, methods to assess potential effects and likelihood of ship strikes to marine mammals in the Marine RSA are lacking. Underwater Noise JASCO’s Marine Operations Noise Model (MONM), used to determine noise sound fields radiating from transiting tankers and escorts, is a state of the art model that incorporates important parameters related to acoustic propagation characteristics in the Marine RSA. Sound levels at various distances from a modelled ship are considered relative to existing acoustic thresholds reported as sound pressure or exposure levels in decibels (dB) associated with auditory injury (Temporary Hearing Threshold Shifts TTS, and Permanent Hearing Threshold Shifts PTS) as well as an acoustic threshold believed to be associated with behavioural disturbance (Southall et al. 2007; NOAA 2013). The resulting outputs are measures of the distance, in kilometres, from the ship at which noise above these thresholds would occur. This review, however, finds that the model has not been applied in a manner that fully assesses ship noise in the Marine RSA. Volume 8A, Marine Transportation, section 4.3 Effects Assessment – Marine Vessel Traffic operations, subsection 4.3.7.4.4 Summary of Acoustic Modelling Results presents four different scenarios with Aframax tankers at different speeds and with different tug escorts. The resulting radii of underwater sound pressure level contours are presented in Table 4.3.7.3. Areas that are known to be within critical habitat for the indicator marine mammals species, for example, the Boundary Pass region, a known area of frequent transit by Southern Resident Killer Whales, have not been included in the four modelling locations. The oceanographic conditions in the Boundary Pass region may vary significantly from the conditions in Haro Strait (Hauser et al. 2007). To adequately assess the exposure of marine mammals to underwater noise, the model should have been run along the whole ship-track, with a realistic combination of tanker

9 154 Pacific Region Trans Mountain Expansion Project

speed and tug escort. With this configuration, a map could have been generated to show the radii of underwater sound pressure level contours along the complete ship track through the Marine RSA, from which the spatial and temporal overlap with known distributions of indicator species could be determined. Data used in the MONM may not be adequate to characterize the acoustic propagation properties of the Marine RSA. In the description of MONM, it is stated that the sound speed profiles used in the modeling come from the U.S. Naval Oceanographic Office’s Generalized Digital Environmental Model (GDEM), with monthly profiles in latitude/longitude grid with 0.25o resolution, which translates to approximately 15 nm between grid points. It is not clear how well these data represent the sound speed profiles in tidally controlled areas like Haro Strait and Boundary Pass. More appropriate data from these areas are available in DFO’s archives. The modeling should also be conducted for typical winter and summer conditions to allow for evaluation of seasonal differences in sound propagation characteristics, and to determine whether certain locations need further attention, especially if they are known to be visited frequently by marine mammals. Volume 8A, Marine Transportation, section 4.3 Effects Assessment – Marine Vessel Traffic operations, subsection, 4.3.7.4.5, Assessment of Potential for Residual Effects of Auditory Injury The assessment of potential harm from the additional shipping activity is primarily focused on the more serious, but less likely, temporary and permanent threshold shifts (TTS and PTS) that could occur in marine mammals exposed to transiting ships. Based on results from the MONM the Proponent concludes that noise-induced temporary threshold shifts and permanent threshold shifts from the increased Project-related shipping are unlikely because noise at such levels would only occur quite close to the ships. However effects from chronic exposure to noise levels that are below the TTS threshold level is a significant concern in the Marine RSA, particularly for SARA listed cetaceans (Erbe et al. 2012; Williams et al. 2013; DFO, 2011). The underwater noise environment in the Marine RSA is not adequately modelled in the Project Application; only Project-related ship noise is modeled, and not the additive and cumulative effects of existing ship source noise. The model is currently used only to assess the impact of a single Project-related ship passing a single stationary marine mammal. As the number of ships transiting the area from all sources increases, the frequency and duration of relative quiet will decrease correspondingly. Model outputs that include additive and cumulative effects of Project- related and existing ship noise would be a more accurate measure of the noise environment to which the marine mammals would be exposed.

Conclusions There are deficiencies in both the assessment of potential effects resulting from ships strikes and exposure to underwater noise in the Trans Mountain Expansion Project Application documents. There is insufficient information and analysis provided with which to assess ship strike risk in the Marine RSA from either existing or Project-related traffic. Ship strike is a threat of conservation concern, particularly for baleen whales such as Fin Whales, Humpback Whales and other baleen whales (Gregr et al. 2006). If shipping intensity increases as projected in Section 4.4 in the Marine RSA and the Strait of Georgia and Juan de Fuca Strait as a whole, the significance of this threat to cetacean populations that occupy the region will increase. Incidence of recovered whale carcasses is not considered to be an adequate measure of the frequency of ship strikes. No information is provided about the speed and maneuverability of Project-related ships or the distribution of whales in relation to the shipping lanes. Analyses that

10 155 Pacific Region Trans Mountain Expansion Project

consider the statistical probability of ship-whale encounters and the risk of collisions are considered appropriate methodologies to assess this potential effect. The JASCO MONM model, as it has been applied by the Proponent, is not adequate to assess the overall impact of noise from increased Project-related traffic. Although state-of-the-art acoustic modelling has been used to model the noise propagation associated with a single Project-related tanker in the Marine RSA, only four locations were chosen to represent the Marine RSA; therefore, the assessment does not adequately represent the noise exposure for the entire time a marine mammal would be in the RSA. The assessment represents only Project-related tanker traffic and not the current noise environment or the potential increase due to Project-related traffic. Finally, the method used to assess the significance of impacts from the modelled noise level contours resulting from a single Project-related tanker and tug on indicator cetacean and pinniped species is qualitative and the lack of an appropriate assessment framework reduces DFO’s ability to evaluate the assessment. Contributors Contributor Affiliation Linda Nichol Author, DFO Science, Pacific Region Svein Vagle Author, DFO Science, Pacific Region Marilyn Hargreaves Author, DFO Science, Pacific Region John Ford Author, DFO Science, Pacific Region Alston Bonamis Author, DFO Fisheries Protection Program, Pacific Region Tola Coopper Editor, DFO Fisheries Protection Program, Pacific Region Lesley MacDougall Editor, DFO Science, Pacific Region

Approved by Carmel Lowe Regional Director Science Branch, Pacific Region Fisheries and Oceans Canada January 14, 2015 Sources of information Clark, C.W., W.T. Ellison, B.L. Southall, L. Hatch, S.M. Van Parijs, A. Frankel and D. Ponirakis. 2009. Acoustic masking in marine ecosystems: Intuitions, analysis, and implication. Marine Ecological Progress Series 395:201–222. Douglas, A.B., J. Calambokidis, S. Raverty, S.J. Jeffries, D.M. Lambourn and S.A. Norman. 2008. Incidence of ship strikes of large whales in Washington State. Journal of the Marine Biological Association of the United Kingdom, 2008, 88(6), 1121–1132. Erbe, C., MacGillivary, A. and R. Williams. 2012. Mapping cumulative noise from shipping to inform marine spatial planning. J. Acoust. Soc. Am. 132(5). Fisheries and Oceans Canada. 2011. Recovery Strategy for the Northern and Southern Resident Killer Whales (Orcinus orca) in Canada. Species at Risk Act Recovery Strategy Series, Fisheries & Oceans Canada, Ottawa, ix + 80 pp.

11 156 Pacific Region Trans Mountain Expansion Project

Gregr, E.J., J. Calambokidis, L. Convey, J.K.B. Ford, R.I. Perry, L. Spaven, M. Zacharias. 2006. Recovery Strategy for Blue, Fin, and Sei Whales (Balaenoptera musculus, B. physalus, and B. borealis) in Pacific Canadian Waters. In Species at Risk Act Recovery Strategy Series. Vancouver: Fisheries and Oceans Canada. vii + 53 pp. Hauser, D.D.W., M.G. Logsdon, E.E. Holmes, G.R. VanBlaricom, R.W. Osborne, 2007, Summer distribution patterns of southern resident killer whales Orcinus orca: core areas and spatial segregation of social groups, Mar. Ecol. Prog. Ser. 351:301-310. Laist, D. W., A. R. Knowlton, J. G. Mead, A. S. Collet, and M. Podesta. 2001. Collisions between ships and whales. Marine Mammal Science 17(1):35-75. Lawson, J.W. and Lesage, V. 2012. A draft framework to quantify and cumulate risks of impacts from large development Projects for marine mammal populations: A case study using shipping associated with the Mary River Iron Mine Project. DFO Can. Sci. Advis. Sec. Res.Doc. 2012/154. iv + 22 p. (Accessed 29 December 2014) MacGillivray, A., G. Warner, and D. Hannay. 2012. Northern Gateway Pipeline Project: Audiogram-Weighted Behavioural Thresholds for Killer Whales. Version 3.0. Technical memorandum by JASCO Applied Sciences for Stantec Consulting Ltd. for Northern Gateway Pipeline Project. National Energy Board. 2013a. Trans Mountain Pipeline ULC – Trans Mountain Expansion: List of Issues. (Accessed 29 December 2014) National Energy Board. 2013b. Filing Requirements Related to the Potential Environmental and Socio-Economic Effects of Increased Marine Shipping Activities, Trans Mountain Expansion Project. (Accessed 29 December 2014) National Oceanic and Atmospheric Administration. 2013 Draft Guidance for Assessing the Effects of Anthropogenic Sound on Marine Mammals. (Accessed December 2014). NRC. 2005. Marine mammal populations and ocean noise. Determining when noise causes biologically significant effects. National Academic Press, Washington, DC. 126pp. (Accessed 29 December, 2014) Rolland, R.M., Parks, S.E., Hunt, K.E., Castellote, M., Corkeron, P.J., Nowacek, D.P., Wasser, S.K. & Kraus, S.D. (2012). Evidence that ship noise increases stress in right whales. Proc. R. Soc. B. Biol Sci 279: 2363–2368. Southall, B. L., A. E. Bowles, W. T. Ellison, J. J. Finneran, R. L. Gentry, C. R. J. Greene, D. Kastak, D. R. Ketten, J. H. Miller, P. E. Nachtigall, W. J. Richardson, J. A. Thomas, and P. L. Tyack (2007). Marine mammal noise exposure criteria: initial scientific recommendations. Aquatic Mammals, Vol. 33(4): 410-522. Vanderlaan, A.S.M., Taggart, C., Serdynska, A.R., Kenney, R.D., & Brown, M. (2008). Reducing the risk of lethal encounters: vessels and right whales in the Bay of Fundy and on the Scotian Shelf. Endangered Species Research, 4: 283-297. doi: 10.3354/esr00083 Williams, R., C.W. Clark, D. Ponirakis and E. Ashe. 2013. Acoustic quality of critical habitats for three threatened whale populations. Animal Conservation 17: 174-185 Williams R. and P. O'Hara. 2010. Modelling ship strike risk to fin, humpback and killer whales in British Columbia, Canada. J. Cetacean Research and Management. 11(1):1– 8.

12 157 Pacific Region Trans Mountain Expansion Project

This Report is Available from the Centre for Science Advice Pacific Region Fisheries and Oceans Canada 3190 Hammond Bay Road Nanaimo, BC V9T 6N7 Telephone: 250-756-7208 E-Mail: [email protected] Internet address: www.dfo-mpo.gc.ca/csas-sccs/ ISSN 1919-3769 © Her Majesty the Queen in Right of Canada, 2015

Correct Citation for this Publication: DFO. 2015. Sufficiency review of the information on effects of underwater noise and the potential for ship strikes from Marine Shipping on Marine Mammals in the Facilities Application for the Trans Mountain Expansion Project. DFO Can. Sci. Advis. Sec. Sci. Resp. 2015/007. Aussi disponible en français : MPO. 2015. Examen du caractère suffisant de l'information sur les effets du transport maritime sur les mammifères marins présentée dans la demande d'installation relative au projet d'agrandissement de Trans Mountain. Secr. can. de consult. sci. du MPO, Rép. des Sci. 2015/007.

13 158

Canadian Science Advisory Secretariat Pacific Region Science Response 2015/022

TECHNICAL REVIEW OF PREDICTED EFFECTS AND PROPOSED MITIGATION OF UNDERWATER NOISE AND POTENTIAL VESSEL STRIKES ON MARINE MAMMALS, FROM THE DECEMBER 2013 FACILITIES APPLICATION AND SUPPLEMENTAL INFORMATION FOR THE TRANS MOUNTAIN EXPANSION PROJECT

Context Trans Mountain Pipeline ULC (Trans Mountain) is proposing an expansion of its current 1,150 km pipeline system between Strathcona County, AB and Burnaby, BC (the Trans Mountain Expansion Project, hereafter ‘the Project’). On December 16, 2013, Trans Mountain filed a Facilities Application for the Project with the National Energy Board (NEB), pursuant to the National Energy Board Act, and on April 2, 2014, the NEB determined that the Application was complete, thereby commencing a 15-month environmental assessment review under the Canadian Environmental Assessment Act, 2012. As part of the application, NEB included a requirement that the Proponent (Trans Mountain) consider the environmental and socio-economic effects of the increase in marine tanker traffic; this was delivered by way of issue # 5 in the NEB’s list of Issues that will be considered at the Hearings (NEB 2013a) and was provided to Trans Mountain in a letter outlining its filing requirements with respect to this topic (NEB 2013b). Through scoping, the Proponent has identified underwater noise from Project-related marine vessel traffic as a potential source of sensory disturbance to marine mammals. They consider ship strikes to marine mammals as a potential accident or malfunction that could occur between marine mammals and Project-related traffic. The assessment of potential effects of the increase in Project-related marine vessel traffic is centered on the established in-bound and out-bound marine shipping lanes in the Marine Regional Study Area (Marine RSA) (Figure 1). As an intervenor in the environmental assessment hearing process for the Trans Mountain Project, Fisheries and Oceans Canada (DFO) will be presenting written evidence to the NEB in relation to its expertise on the effects of the Project on marine fish and fish habitat and marine mammals (including aquatic species at risk), the efficacy and adequacy of mitigation and offsetting measures, monitoring and follow-up programs proposed by the Proponent, and the conclusions reached in Facilities Application for the Project. DFO’s Pacific Region Fisheries Protection Program (FPP) is responsible for reviewing potential effects of the marine terminal and shipping components of the Project on fish, fish habitat and marine mammals. In December 2014, FPP requested that DFO Science Branch conduct a sufficiency review of the information on the effects of marine shipping on marine mammals in the Facilities Application for the Project, and in February 2015, DFO Science Branch published a Canadian Science Advisory Secretariat (CSAS) Science Response (SR) to this request. This CSAS SR contributed to DFO’s information request (DFO IR No.2) filed with the NEB in

May 2015

159 Science Response: Technical Review Trans Pacific Region Mountain Expansion

Round 2 of information requests made of the Proponent. Trans Mountain responded to DFO IR No.2 on February 13, 2015. Based on information provided by the Proponent within the Facilities Application and in relevant supplemental filings with the NEB (including Trans Mountain’s response to DFO IR No. 2), FPP is now requesting DFO Science Branch provide advice on whether the Proponent’s assessment of effects related to underwater noise and ship strikes on marine mammals from increased Project-related vessel traffic supports their conclusions. Furthermore, FPP is seeking advice to understand if effects related to underwater noise and ship strikes on marine mammals are measureable in relation to current conditions, and whether there are mitigation measures available to reduce these effects. DFO Science Branch has been asked to focus its review on direct effects on marine mammals, not including potential effects of underwater noise on prey species, such as fish. The assessment and advice arising from this CSAS SR will assist in the development of DFO’s written evidence submission to the National Energy Board for the Trans Mountain Expansion Project. The objective of this SR is to review information provided by the Proponent in the December 2013 Facilities Application and in relevant supplemental filings with the NEB to answer the following questions: 1. Does the Proponent’s assessment of the effects of underwater noise and ship strikes on marine mammal indicator species (i.e., the Southern Resident Killer Whale, Humpback Whale, and Steller Sea Lion) accurately characterize potential direct effects on these marine mammal species within the Marine Regional Study Area1 (Marine RSA), and does their assessment support their conclusions? 2. For this project, is it possible to measure the effect of underwater noise and mammal- vessel ship strikes from increased Project-related vessel traffic on marine mammal indicator species above the status quo (i.e., current noise conditions), and if so, how does this effect differ from current conditions? 3. If measureable effects exists in (2) above, are there mitigation measures that may be implemented to reduce the residual effect on the marine mammals indicator species? This Science Response Report results from the Science Response Process of March 20, 2015 on the Technical review of predicted effects and proposed mitigation of underwater noise and potential vessel strikes on marine mammals, from the December 2013 Facilities Application and supplemental information for the Trans Mountain Expansion Project.

2 160 Science Response: Technical Review Trans Pacific Region Mountain Expansion

3 161 Science Response: Technical Review Trans Pacific Region Mountain Expansion

Background The existing Trans Mountain pipeline (TMPL) system commenced operation in 1953, and transports a range of crude oil and petroleum products from Western Canada to locations in central and southwestern British Columbia (BC), Washington State, and offshore. The proposed Project would create a twinned pipeline, increasing the capacity of the system from approximately 300,000 barrels per day to 890,000 barrels per day. Key project components include 994 km of new pipeline, reactivation of 193 km of existing pipeline, 12 new pump stations and expansion of existing pump stations and storage tanks, and the addition of three new vessel berths at the Westridge Marine Terminal in Burnaby, BC. For this particular Science Branch review, the marine vessel traffic that will transport the petroleum products is of relevance. The proposed expansion is forecasted to increase marine vessel traffic from 5 tankers per month calling at the Westridge Marine Terminal to approximately 34 tankers per month (i.e., an additional 720 tanker transits each year). At present, the maximum size of petroleum tankers that call at the Westridge Terminal are Aframax class, which have an average cargo carrying capacity of 750,000 barrels. The maximum size of tankers is not expected to change as part of the Project. These vessels will transit the Marine Regional Study Area (Marine RSA) using existing in-bound and out-bound shipping lanes (Figure 1). It will take each Project-related Marine vessel approximately 12 hours to complete one transit of the Marine RSA; and, on average, there will be two transits every 24 hours. This will be in addition to existing traffic in the shipping lanes and other traffic in the Marine RSA. There are 22 species of marine mammals identified by the Proponent as being present in the Marine RSA. The most commonly observed species of toothed whales in the Marine RSA include Killer Whales (Orcinus orca), Harbour Porpoises (Phocoena phocoena), Dall’s Porpoises (Phocoenoides dalli), and Pacific White-sided Dolphins (Lagenorhynchus obliquidens). Designated Critical Habitat for the Southern Resident Killer Whale population listed as endangered under the Species at Risk Act (SARA) overlaps almost entirely with the Marine RSA (Figure 2). The Humpback Whale (Megaptera novaeangliae), which is listed as threatened under the SARA, is the most commonly observed baleen whale, and the western- most portion of the Marine RSA overlaps proposed Humpback Whale Critical Habitat (Figure 2). Other baleen whales including the Minke Whale (Balaenoptera acutorostrata) and Grey Whale (Eschrichtius robustus), as well as the occasional Fin Whale (Balaenoptera physalus) are also observed. The Harbour Seal (Phoca vitulina), and Steller (Eumetopias jubatus) and California (Zalophus californianus) Sea Lions are the most common pinnipeds observed in the Marine RSA, as well as the occasional Northern Elephant Seal (Mirounga angustirostris). Sea Otters (Enhydra lutris) are also occasionally sighted. The Proponent’s assessment of potential effects is on three marine mammal species identified as indicators of toothed whales, baleen whales and pinnipeds in the Marine RSA; respectively these are Southern Resident Killer Whales, Humpback Whales, and Steller Sea Lions. An increase in marine vessel traffic associated with the Project has the potential to result in sensory disturbance to marine mammals from underwater noise and an increased risk of injury and mortality associated with mammal-vessel strikes. Disturbance responses associated with increased Project-related vessel traffic could range from temporary displacement to reduced foraging efficiency, to disruption of mating and social behaviours. The potential for these effects to affect recovery of the Southern Resident Killer Whale is of critical importance, as fewer than 80 individuals are estimated to be present in the wild. Furthermore, the Proponent has noted in the Project Application that although the Project’s contribution to overall sensory disturbance effects on the species is small, the potential effects of increased Project-related marine vessel traffic are determined to be significant for Southern Resident Killer Whales. This species is therefore of the greatest conservation concern in the Marine RSA.

4 162 Science Response: Technical Review Trans Pacific Region Mountain Expansion

Figure 2. Critical Habitat for Southern Resident Killer Whales, proposed Critical Habitat for Humpback Whales and other important areas for marine mammals in the Marine RSA (from Trans Mountain Pipeline ULC, 2013. Trans Mountain Expansion Project – An Application Pursuant to Section 52 of the National Energy Board Act, Volume 8A - Marine Transportation).

5 163 Science Response: Technical Review Trans Pacific Region Mountain Expansion

Analysis and Response Science Branch responses to the three technical assessment questions posed by DFO FPP (see Context above) are detailed below. To prepare this response, the following documents from the Proponent were reviewed: 1. Trans Mountain Pipeline ULC. 2013. Trans Mountain Expansion Project – An Application Pursuant to Section 52 of the National Energy Board Act, Volume 8A - Marine Transportation. Submitted to the Secretary of The National Energy Board. 2. JASCO Applied Sciences. 2014. Supplemental Underwater Noise Modelling for Trans Mountain Expansion Project. Prepared for Stantec Ltd. Document 00542, Version 5.0: 51 p. 3. Trans Mountain Pipeline ULC. 2014. Trans Mountain Response to The National Energy Board Information Request No. 1. NEB Hearing Order OH-001-2014. Trans Mountain Document Number: A3W9H8. Submitted to the Secretary of The National Energy Board. 4. Trans Mountain Pipeline ULC. 2014. Trans Mountain Response to Information Request from Val Veirs. NEB Hearing Order OH-001-2014. Trans Mountain Document Number: A3X6V7. Submitted to the Secretary of The National Energy Board. 5. Moffat & Nichol. 2013. TERMPOL 3.2 – Origin, Destination & Marine Traffic Volume Survey. Trans Mountain Expansion Project. Prepared for Trans Mountain. Trans Mountain Document Number: A3S4R7 and A3S4R8. Submitted to the Secretary of The National Energy Board. 6. Trans Mountain Pipeline ULC. 2014. Trans Mountain Response to The National Energy Board Information Request No. 2. NEB Hearing Order OH-001-2014. Trans Mountain Document Number: A3Z4T9. Submitted to the Secretary of The National Energy Board. 7. Fisheries and Oceans Canada. 2015. Sufficiency review of the information on effects of underwater noise and the potential for ship strikes from marine shipping on marine mammals in the facilities application for the Trans Mountain expansion Project. DFO Can. Sci. Advis. Sec. Sci. Resp. 2015/007. 8. Trans Mountain Pipeline ULC. 2015. Trans Mountain Response to Government of Canada Information Request No. 2. NEB Hearing Order OH-001-2014. Trans Mountain Document Number: A4H6A5. Submitted to the Secretary of The National Energy Board.

Does the Proponent’s assessment of the effects of underwater noise and ship strikes on marine mammal indicator species (i.e., the Southern Resident Killer Whale, Humpback Whale, and Steller Sea Lion) accurately characterize potential direct effects on these marine mammal species within the Marine Regional Study Area (Marine RSA), and does their assessment support their conclusions? Underwater Noise The Proponent has provided a comprehensive review of the current literature on underwater noise and its potential effects on marine mammals, with particular emphasis on the three indicator species; the Southern Resident Killer Whale, Humpback Whale, and Steller Sea Lion. In the absence of Canadian standards or guidelines for exposure of marine mammals to chronic (i.e., non-pulsed) underwater anthropogenic noise, the Proponent has used thresholds for auditory injury (Temporary Hearing Threshold Shifts TTS, and Permanent Hearing Threshold Shifts PTS) recommended by Southall et al. (2007) and thresholds for behavioural disturbance currently used in the US (NOAA 2013). These are commonly applied standards that have also been used by DFO in the past (e.g. Lawson and Lesage 2012). The Proponent has also

6 164 Science Response: Technical Review Trans Pacific Region Mountain Expansion provided a detailed assessment of the anticipated noise sound fields radiating from transiting Project-related tankers and escorts, based on a state-of-the-art model that incorporates important parameters related to acoustic propagation characteristics in the Marine RSA (JASCO’s Marine Operations Noise Model (MONM)). Modelled sound levels at various distances from a theoretical ship and escort tug were considered relative to existing acoustic thresholds, reported as sound pressure or exposure levels in decibels (dB) associated with auditory injury as well as potential behavioural disturbance (Southall et al. 2007; NOAA 2013). The resulting outputs are measures of the distance, in kilometres, from the ship at which noise above these thresholds would occur in various scenarios of vessel configuration and speed. As noted in DFO’s sufficiency review of the Project Application (DFO 2015), this model, while exemplary, was not applied in a manner that fully assesses ship noise in the Marine RSA, and thus the assessment of the potential effects of Project-related noise on indicator species is incomplete. In particular, three areas of deficiency were identified in this review. First, the modelling exercise only considered predicted noise properties at four locations along the 296 km long transit corridor (shipping lanes in Figure 1) for Project-vessels in the Marine RSA. While these locations may be reasonably representative of conditions over much of the Marine RSA, they do not necessarily reflect acoustic propagation properties that might occur in some important areas within the designated Critical Habitat of Southern Resident Killer Whales. Second, the dataset originally used in the MONM assessment is inferior to other available datasets (DFO Institute of Ocean Sciences Data Archive. Ocean Sciences Division), in terms of seasonal sound speed profiles. Third, the MONM was used to assess the impact of a single Project-related ship with an escort tug passing a single stationary marine mammal. A complete assessment of the impacts of this Project would require that the model be used to estimate the frequency with which such events would occur from all shipping in the Marine RSA to a theoretical stationary marine mammal. This is necessary to assess chronic and additive noise levels in the context of shipping activity associated with the Project. Model outputs that include additive and cumulative effects of Project-related and existing ship noise would be a more accurate measure of the noise environment to which the marine mammals would be exposed. Specifically, the impacts of additional ship noise associated with Project-related vessels on periods of relative quiet should be modeled in addition to existing shipping in the Marine RSA. In response to DFO’s requests that these deficiencies be addressed (Trans Mountain Pipeline ULC 2015), the Proponent stated that the four locations used in the MONM were sufficiently representative of other areas within the Marine RSA, and any differences would be unlikely to change their conclusions with respect to impacts of underwater vessel noise on marine mammals (Response to IR No. 2.088 in Trans Mountain Pipeline ULC 2015). In addition, the Proponent maintained that modeling of noise from existing shipping in the Marine RSA was not possible due to a lack of data on current ambient noise in the region (Response to IR No. 2.087 in Trans Mountain Pipeline ULC 2015). However, there is a considerable amount of ambient noise data available for the Strait of Georgia from hydrophones on the VENUS cabled observatory (Ocean Networks Canada). There are also considerable AIS (Automatic Identification System) data on transits of all types of commercial shipping vessels in the Marine RSA in recent years (e.g., Simard et al. 2014) and typical source sound pressure levels of noise generated by vessel type are readily available (e.g., McKenna et al. 2012). A recent coast-wide assessment of shipping noise in BC using AIS data was recently published by Erbe et al. (2012 and 2014). The Proponent’s conclusions with regard to impacts of underwater noise due to Project-related vessel traffic on marine mammals in the Marine RSA are presented in Volume 8 of the Project Application, Section 4.3.7.5 Potential Residual Effects and Table 4.3.7.8. They conclude that the only residual effect of increased Project-related marine vessel traffic on all marine mammals indicator species includes Sensory disturbance due to underwater noise from vessels (may

7 165 Science Response: Technical Review Trans Pacific Region Mountain Expansion include temporary displacement, startle response, increased energy expenditure, reduced foraging efficiency, communication masking, change in activity state, and/or increased stress). The significance of this residual effect for each of the three indicator species is evaluated in Section 4.3.7.6 and summarized in Table 4.3.7.9. For Southern Resident Killer Whales, the Project Application concludes that there would be a significant residual environmental effect, which is defined as “a high probability of occurrence of a permanent or long-term residual effect of high magnitude that cannot be technically or economically mitigated”. The criteria used to reach this conclusion include the duration of the disturbance (long-term), its frequency of occurrence (periodic), its reversibility (or time to recover from a disturbance event – immediate), magnitude of the disturbance (high), its probability of occurrence (high), and confidence that the assessment of significance is correct (low). Of the various reasons given for the low confidence in this significance assessment, the “greatest source of uncertainty is the linkage of sensory disturbance effects to population-level consequences and the degree to which such effects can be attributed to underwater noise from Project-related marine vessels and other ships and boats”. For Humpback Whales and Steller Sea Lions, any residual effect is deemed to be not significant. Overall, while the Proponent’s assessment is largely qualitative and has a low level of confidence, their conclusion that there would be a significant residual effect from Project-related vessel noise on Southern Resident Killer Whales is still consistent with DFO Science advice; given the current state of knowledge and declining trend of this population in recent decades. However, it is important to note that their assessment is based only on Project-related activities, and does not consider the potential cumulative and/or additive effects of noise generated by Project-related vessels on top of existing shipping. Thus, sound exposure levels would likely be in excess of those estimated in their assessment, as would the proportion of time that whales would be exposed to potentially disturbing levels of shipping noise. Of the criteria used to determine significance of residual effects, their conclusion that reversibility would be immediate, i.e., within two days, appears unrealistic. Based on the Proponent’s predicted two Project- related ship transits through the Marine RSA per day, amounting to exposures of potentially disturbing noise levels (>120 dB re: 1 µPa) for up to 8% of the day, such immediate reversibility may not occur. The Proponent concludes that potential residual effects of Project-related vessel noise on Humpback Whales would not be significant. The rationale for this conclusion appears to be based mostly on the anticipated low numbers of individuals exposed relative to the population abundance of Humpback Whales in BC and the North Pacific generally. However, their assessment fails to take into consideration the strong long-term site fidelity exhibited by individual Humpback Whales to particular feeding areas (e.g., Rambeau 2008; Ford et al. 2009). The Marine RSA overlaps with a significant concentration area for Humpback Whales at the western entrance to Juan de Fuca Strait, which was recently identified as potential Critical Habitat (Nichol et al. 2009). Furthermore, more than 100 individual Humpback Whales have been photo-identified in Juan de Fuca Strait south and west of Victoria, and a substantial proportion of these have been observed to use these waters in multiple years (Malleson et al. 2014). Because of potentially high densities of individuals with strong site fidelity, Humpback Whales using the Marine RSA have the potential for repeated exposure to Project-related shipping noise at levels that could potentially cause behavioural disturbance. Thus, the residual effect on humpback whales, from underwater noise generated by Project-related vessel traffic, may be greater than the Proponent has identified, and should be taken into consideration in the determination of significant residual adverse effects. Based on DFO’s review, the residual effect of underwater noise from increased Project-related marine vessel traffic on Steller Sea Lions has been accurately characterized in the Proponent’s

8 166 Science Response: Technical Review Trans Pacific Region Mountain Expansion assessment, and their assessment supports their conclusion that potential residual effects would be negligible. Ship Strike As discussed in detail in DFO’s sufficiency review of the Project Application (DFO 2015), the Proponent considered ship strike risk only in the context of “Accidents and Malfunctions” and not as a result of routine operations. As such, a detailed quantitative assessment was not undertaken of potential residual effects of ship strikes on marine mammals from Project-related vessel traffic. Although their qualitative conclusion that risk to Southern Resident Killer Whales and Steller Sea Lions from Project-related vessel collision may be reasonable (extremely low or negligible), this may not be the case for Humpback Whales. There is ample scientific evidence that Humpback Whales are regularly struck and injured or killed by large commercial vessels, including oil tankers (e.g., IWC 2014). Despite the rarity of documented ship strike events in BC waters and the Marine RSA in particular, such records are known to underestimate their true frequency of occurrence. Because of the appreciably higher density of Humpback Whales in Juan de Fuca Strait and its western entrance (within the Marine RSA) than was suggested by the Proponent, the risk to individuals of this species could be greater than portrayed in their assessment. Thus, it is likely that the Proponent’s assessment of potential direct effects on Humpback Whales from Project-related vessel collision is inaccurate.

For this project, is it possible to measure the effect of underwater noise and mammal-vessel ship strikes from increased Project-related vessel traffic on marine mammal indicator species above the status quo (i.e., current noise conditions), and if so, how does this effect differ from current conditions? Underwater Noise A comprehensive assessment of the potential effects of increased Project-related vessel noise and the cumulative and additive effects of this increased noise in combination with existing vessel-related noise in the Marine RSA is possible but has not yet been undertaken. Such an assessment would require: 1. detailed characterization of existing ambient noise generated by marine traffic throughout the Marine RSA and expanded modelling of the noise propagation properties over the transit corridor; and 2. improved understanding of the effects of increased vessel noise on the behaviour and population dynamics of the marine mammal species of concern (primarily Southern Resident Killer Whales). Noise characterization As described above, an assessment of existing shipping traffic noise in the Marine RSA is possible, but was not undertaken by the Proponent. Sufficiently high-resolution AIS data for the region, including vessel type and speed, are available to quantify vessel activity throughout the year. From these data, estimated noise levels could be derived using nominal source sound pressure level data available for the various types of ships involved. These estimates would be preliminary, but could be improved by analysis of existing data on ambient noise in the Marine RSA that have been collected by DFO (DFO Cetacean Research Program; unpublished data), the VENUS project (Ocean Networks Canada) and Williams et al. (2013). Such an analysis is currently being conducted as part of the Marine Environmental Observation Prediction and Response Network (MEOPAR) funded project titled “Modeling Ship Movements: Application for Noise Exposure to the Marine Ecosystem”, a collaboration among researchers with the University of Victoria, DFO and others. Similar efforts that are being promoted as part of Port

9 167 Science Response: Technical Review Trans Pacific Region Mountain Expansion

Metro Vancouver’s Enhancing Cetacean Habitat and Observation (ECHO) Program, and through a new Whale Tracking Network project involving DFO, will add considerably to the collection of ambient acoustic data for the Marine RSA. With baseline data on existing noise levels in the Marine RSA, additional noise levels due to Project-related vessels could be modelled to determine cumulative and/or additive levels of exposure of acoustic disturbance on marine mammals. Effects of Noise on Marine Mammals Details on the effects of chronic industrial noise on marine mammals, including that emanating from shipping, are not well understood. Existing knowledge of the short-term responses of Killer Whales and Humpback Whales to anthropogenic noise is described in the Proponent’s impact assessment. Field studies on Resident Killer Whales in BC and Washington State suggest that behavioural disturbance due to vessels and accompanying noise has the potential to disrupt life processes such as communication, echolocation and foraging. However, the long-term consequences of such disturbance in terms of its effects on population dynamics (e.g., mortality and birth rates) are not known. The Recovery Strategy for Resident Killer Whales (Fisheries and Oceans Canada 2011b) identifies several threats in addition to anthropogenic noise, such as bioaccumulating persistent organic pollutants (POPs) and food supply, and these may act synergistically. The poor survival and birth rates of Southern Resident Killer Whales over the past 20 years suggest that current habitat quality, including that of designated Critical Habitat within the Marine RSA, is insufficient to allow for recovery of this population. Although ambient noise from shipping is only one of several anthropogenic stressors, existing noise levels may already be having a significant effect on population survival. However, in the absence of information needed to quantify the impact of existing noise conditions in the Marine RSA on Southern Resident Killer Whales population dynamics, it is not possible to predict what further effects might be anticipated from future Project-related vessel traffic. Recent efforts to develop frameworks to quantify the cumulative effects of noise and other stressors on population dynamics of marine mammals (e.g., the Population Consequences of Acoustic Disturbance (PCAD) model, NRC 2005; Lawson and Lesage 2012 may provide a means of measuring such future impacts. Ship strike It is possible to estimate the current risk to marine mammal indicator species from ship strikes in the Marine RSA and the additional risk that could result from Project-related vessel traffic, but such estimates would likely have a high degree of uncertainty. As described previously, quantifying the spatial and temporal distribution of existing ship movements through the Marine RSA can be accomplished using archived AIS data that are readily available (e.g., Simard et al. 2014). However, high resolution spatial data on the densities of marine mammal indicator species are lacking, particularly for the Humpback Whale, which is the only species that is at significant risk of collision. Thus, an accurate quantitative assessment of risk to Humpback Whales from existing shipping traffic is not feasible at this time, nor is an estimate of the increased risk associated with Project-related vessels.

If measureable effects exist in (2) above, are there mitigation measures that may be implemented to reduce the residual effect on the marine mammals indicator species? Underwater Noise Underwater noise generated by commercial shipping vessels and their effects on marine mammals is of growing concern internationally. Recently, there have been increasing efforts to characterize in detail the noise spectrum and levels from modern shipping (e.g., McKenna et al.

10 168 Science Response: Technical Review Trans Pacific Region Mountain Expansion

2013), and to develop means of reducing the noise output of ships through propeller and ship design (e.g., RMC 2009, unpublished report1). There are a number of initiatives, both internationally and regionally, that focus on anthropogenic ocean noise and its mitigation, including noise caused by marine shipping. For example, the European Union (EU) has initiated a multinational collaborative project called SONIC (Suppression Of underwater Noise Induced by Cavitation) which aims to develop tools to investigate and mitigate the effects of underwater noise generated by shipping. Another EU-supported initiative is the AQUO (Achieve QUieter Oceans) project, a consortium of European government, industry and academic organizations with the objective of assessing and mitigating noise impacts of the maritime transportation on the marine environment. The final goal of this project is to provide policy makers with practical guidelines, acceptable to shipyards and ship owners, needed to mitigate underwater noise footprints due to shipping and to prevent negative consequences to marine life. In the US, the National Oceanic and Atmospheric Administration (NOAA) has recently developed an Ocean Noise Strategy, a 10-year program to develop management actions to reduce the acute, chronic, and cumulative effects of noise. To date, DFO has not endorsed nor implemented any particular shipping noise management program. In Pacific Region, there are several initiatives underway to document underwater shipping noise and to develop means of mitigating this noise. World Wildlife Fund (WWF) Canada has recently convened a workshop “to discuss various methods for minimizing and mitigating underwater noise, and to develop tools for planners and regulators to draw from as they move forward in planning processes, reviewing environmental assessments, and recovery planning for species listed under the Species at Risk Act (SARA)”. The workshop report (WWF-Canada 2013) describes the various approaches that are currently being developed in different jurisdictions to mitigate noise from shipping. Port Metro Vancouver (PMV) is also assessing potential options for noise mitigation through its ECHO program. One initiative is a pilot study that Transport Canada is planning that involves the use of All-Sea Whale Shark hull cleaning technology within PMV jurisdiction. The pilot will review what impact hull cleaning has on both fuel efficiency and underwater noise outputs. DFO has participated in working groups to identify data gaps and proposed studies to address the gaps, but a national ocean noise strategy does not currently exist. In summary, there are potentially a variety of different methods by which increased exposure of marine mammals to noise by Project-related vessels in the Marine RSA could be mitigated. Ship Strike There has been considerable international effort in recent years to quantify the risk of ship strikes to various species of whales in different regions, and to develop methods of mitigation (e.g., Vanderlaan and Taggart 2007; Redfern et al. 2013, Silber et al. 2012; van der Hoop et al. 2013). Mitigation of ship strike risk to whales is achieved primarily through reduction in vessel speed (to 10 knots or less) and/or by altering shipping corridors to avoid concentrations of animals. Both methods have been proven effective in reducing ship strikes on endangered North Atlantic Right Whales. In 2007, the International Maritime Organization adopted the Roseway Basin “Area to Be Avoided” on the Scotian Shelf as a voluntary conservation initiative and this was implemented by Canada (Vanderlaan and Taggart 2009). This initiative led to a significant decrease in the risk of strikes to Right Whales. In 2008, NOAA issued mandatory time-area vessel speed restrictions along the US eastern seaboard, which also significantly reduced the risk of lethal strikes to Right Whales (Conn and Silber 2013). In 2013, the

1 RMC (Renilson Marine Consulting Pty. Ltd.) 2009. Reducing Underwater Noise Pollution from Large Commercial Vessels. Unpubl. Report to International Fund for Animal Welfare.

11 169 Science Response: Technical Review Trans Pacific Region Mountain Expansion

Canadian Coast Guard, in collaboration with Parks Canada and other stakeholders, introduced voluntary seasonal speed restrictions for shipping vessels in the Saguenay-St. Lawrence Marine Park, and also recommended an area to be avoided for Blue Whales (Canadian Coast Guard 2013). In summary, there are potential options for ship strike mitigation that could be applied in the Marine RSA to reduce the risk of mammal-vessel collisions. The Proponent has recognized that alterations in ship speed and routing can be effective mitigation measures to reduce ship strike risk, but has not proposed to adopt such measures (Trans Mountain Response to NEB IR No. 2, pg 254). The Marine Mammal Regulations, under the Fisheries Act, specifically prohibit the disturbance of whales and other marine mammals; however there are currently no regulatory or policy frameworks specific to mitigation of potential ship strike impacts on marine mammals within the Marine RSA.

Conclusions The information provided by the Proponent in the Project Application, and in subsequent responses to information requests by DFO, was determined to have insufficiencies. As a result, DFO’s ability to definitively conclude if the Proponent has accurately assessed direct effects of underwater noise or ship strikes from Project-related vessel traffic on marine mammals is reduced. Based on the information available, DFO’s review of the Project Application and subsequent responses has yielded the following conclusions.

Underwater Noise The Proponent’s assessment likely does not accurately characterize the effects of underwater noise on certain marine mammal indicator species in the Marine RSA. Despite the use of a state-of-the-art noise model, only four sites were modelled along the 296 km shipping corridor, and these sites may not be representative of acoustic properties throughout all portions of the shipping corridor. In addition, model inputs on oceanographic conditions used in the assessment could be improved. Furthermore, the assessment only addressed the noise produced and propagated from Project-related ships, and did not consider the potential cumulative and/or additive effects of such noise in combination with noise produced by existing shipping activity. The Proponent’s conclusion that there would be a significant residual effect from Project-related vessel noise on Southern Resident Killer Whales is plausible; but the potential impact of such an effect on the long-term population viability of the marine mammal indicator species within the Marine RSA is unknown. It is likely that the Proponent has underestimated the potential residual effects of underwater noise on Humpback Whales in the Marine RSA. There are greater seasonal densities of Humpback Whales in the western portions of the Marine RSA than the Proponent has noted, and the strong site fidelity to specific feeding grounds for this species is likely to increase the exposure of individuals to potentially disturbing noise levels from increased Project-related vessel traffic. There are a variety of initiatives underway, both locally and globally, to develop mitigation measures to reduce levels of underwater noise generated by commercial shipping. However, there are currently no specific measures that have been implemented or required within the Marine RSA for mitigation of underwater noise generated by vessels.

Ship Strike The Proponent has not considered ship strike risk to marine mammals in the context of routine operations of Project-related marine vessels. As such, a detailed quantitative assessment of

12 170 Science Response: Technical Review Trans Pacific Region Mountain Expansion potential residual effects of ship strikes on marine mammals from Project-related vessel traffic was not undertaken. Although the risk to Killer Whales and Steller Sea Lions from collisions with large ships is negligible, this may not be the case for Humpback Whales. Humpback Whales can be vulnerable to strikes from large ships including tankers, and there are higher seasonal densities of this species in the western portions of the Marine RSA than suggested by the Proponent. As a result, the Proponent’s assessment of potential direct effects on Humpback Whales from collisions with Project-related vessels is likely an underestimate. There are currently no specific ship-strike mitigation guidelines or measures that have been implemented in the area of the Marine RSA or in Pacific Canadian waters. The Proponent has recognized that alterations in ship speed and routing can be effective mitigation measures to reduce ship strike risk, but has not proposed to adopt such measures because it does not own or operate the shipping vessels, or possess the regulatory authority to require such alterations.

Contributors

Contributor Affiliation

John Ford Author, DFO Science, Pacific Region

Linda Nichol Author, DFO Science, Pacific Region

Svein Vagle Author, DFO Science, Pacific Region

Alston Bonamis Author, DFO Fisheries Protection Program, Pacific Region

Tola Coopper Editor, DFO Fisheries Protection Program, Pacific Region

Marilyn Hargreaves Editor, DFO Science, Pacific Region

Lesley MacDougall Editor, DFO Science, Pacific Region

Approved by Carmel Lowe Regional Director Science Branch, Pacific Region Fisheries and Oceans Canada April 8, 2015

Sources of information Canadian Coast Guard. 2013. Notice to Mariners Eastern Edition. VOL. 38, MONTHLY EDITION NO 05 May 31, 2013 Publication Number 40063779 Conn, P. B., and Silber, G. K. 2013. Vessel speed restrictions reduce risk of collision-related mortality for North Atlantic right whales. Ecosphere, 4(4), art43. DFO. 2015. Sufficiency review of the information on effects of underwater noise and the potential for ship strikes from Marine Shipping on Marine Mammals in the Facilities Application for the Trans Mountain Expansion Project. DFO Can. Sci. Advis. Sec. Sci. Resp. 2015/007. (Accessed 08 Apr 2015)

13 171 Science Response: Technical Review Trans Pacific Region Mountain Expansion

Erbe, C., MacGillivary, A. and R. Williams. 2012. Mapping cumulative noise from shipping to inform marine spatial planning. J. Acoust. Soc. Am. 132(5):EL423-EL428. Erbe, C, Williams, R, Sandilands, D, Ashe, E 2014. Identifying Modeled Ship Noise Hotspots for Marine Mammals of Canada's Pacific Region. PLoS ONE 9(3): e89820. doi:10.1371/journal.pone.0089820 Fisheries and Oceans Canada. Pacific Region. Institute of Ocean Sciences Data Archive. Ocean Sciences Division. [Internet]. (Accessed 08 Apr 2015) Fisheries and Oceans Canada. 2011a. Marine Mammal Regulations. Last Amended 02-10- 2011. (Accessed 08 Apr 2015) Fisheries and Oceans Canada. 2011b. Recovery Strategy for the Northern and Southern Resident Killer Whales (Orcinus orca) in Canada. Species at Risk Act Recovery Strategy Series, Fisheries & Oceans Canada, Ottawa, ix + 80 pp. (Accessed 08 Apr 2015) Ford J.K.B., Rambeau A.L., Abernethy R.M., Boogaards M.D., Nichol L.M., and Spaven L.D. 2009. An Assessment of the Potential for Recovery of Humpback Whales off the Pacific Coast of Canada. DFO Can. Sci. Advis. Sec. Res. Doc. 2009/015. iv + 33 p. (Accessed 08 Apr 2015) Lawson, J.W. and Lesage, V. 2012. A draft framework to quantify and cumulate risks of impacts from large development Projects for marine mammal populations: A case study using shipping associated with the Mary River Iron Mine Project. DFO Can. Sci. Advis. Sec. Res.Doc. 2012/154. iv + 22 p. (Accessed 08 Apr 2015) Malleson, M., E. Heydenreich, K. C. Balcomb III, and D. Ellifrit 2014. Humpback Whales of the island waters of Washington State and Southern British Columbia. Center for Whale Research, Friday Harbor, WA. (Accessed 08 Apr 2015) McKenna, M.F., D. Ross, S.M. Wiggins, J.A. Hildebrand. 2012. Underwater radiated noise from modern commercial ships. J. Acoust. Soc. Am. 131 (1), January 2012. Pp92-103. National Oceanic and Atmospheric Administration (NOAA). 2013 Draft Guidance for Assessing the Effects of Anthropogenic Sound on Marine Mammals. (Accessed 08 Apr 2015). Rambeau, A.L. 2008. Determining abundance and stock structure for a widespread migratory animals: the case of humpback whales (Megaptera novaeangliae) in British Columbia, Canada. M.Sc Thesis. University of British Columbia. Redfern, J. V., McKenna, M. F., Moore, T. J., Calambokidis, J., Deangelis, M. L., Becker, E. A., Barlow, J., Forney, K.A., Fiedler, P.C., and S. J. Chivers. (2013). Assessing the risk of ships striking large whales in marine spatial planning. Conservation Biology, 27(2), 292- 302. Silber, G.K., Vanderlaan, A.S.M. and Tejedor, Arceredillo, A., et al. 2012. The role of the International Maritime Organization in reducing vessel threat to whales: process, options, action and effectiveness. Mar. Policy, 36, 1221-1233. Simard, Y., Roy, N., Giard, S., and Yayla, M. 2014. Canadian year-round shipping traffic atlas for 2013: Volume 3, West Coast. Can. Tech. Rep. Fish. Aquat. Sci. 3091(Vol.3)E: xviii + 327 pp. Southall, B. L., A. E. Bowles, W. T. Ellison, J. J. Finneran, R. L. Gentry, C. R. J. Greene, D. Kastak, D. R. Ketten, J. H. Miller, P. E. Nachtigall, W. J. Richardson, J. A. Thomas, and P. L. Tyack (2007). Marine mammal noise exposure criteria: initial scientific recommendations. Aquatic Mammals, Vol. 33(4): 410-522.

14 172 Science Response: Technical Review Trans Pacific Region Mountain Expansion

Vanderlaan, A.S.M. and Taggart, C.T. 2007. Vessel Collisions with whales: the probability of lethal injury based on vessel speed. Mar. Mammal Sci., 23, 144-156. Vanderlaan, A.S.M. and Taggart, C.T. 2009. Efficacy of a voluntary area to be avoided to reduce risk of lethal vessel strikes to endangered whales. Conservation Biology, 23: 1467–1474. Williams, R., C.W. Clark, D. Ponirakis and E. Ashe. 2013. Acoustic quality of critical habitats for three threatened whale populations. Animal Conservation 17: 174-185 WWF-Canada. 2013. Finding Management Solutions for Underwater Noise in Canada’s Pacific. Vancouver Aquarium and WWF-Canada, Vancouver, B.C. (Accessed 08 Apr 2015)

This Report is Available from the Centre for Science Advice Pacific Region Fisheries and Oceans Canada 3190 Hammond Bay Road Nanaimo, BC V9T 6N7 Telephone: 250 756-7208 E-Mail: [email protected] Internet address: www.dfo-mpo.gc.ca/csas-sccs/ ISSN 1919-3769 © Her Majesty the Queen in Right of Canada, 2015

Correct Citation for this Publication: DFO. 2015. Technical review of predicted effects and proposed mitigation of underwater noise and potential vessel strikes on marine mammals, from the December 2013 Facilities Application and supplemental information for the Trans Mountain Expansion Project. DFO Can. Sci. Advis. Sec. Sci. Resp. 2015/022. Aussi disponible en français : MPO. 2015. Examen technique des effets prévus et des mesures d'atténuation proposées du bruit sous-marin et des collisions avec des navires sur les mammifères marins présentés dans la demande d'installation de décembre 2013 relative au projet d'agrandissement de Trans Mountain et dans ses documents connexes. Secr. can. de consult. sci. du MPO, Rép. des Sci. 2015/022.

15 173

Canadian Science Advisory Secretariat Pacific Region Science Advisory Report 2016/042

INTEGRATED BIOLOGICAL STATUS OF SOUTHERN BRITISH COLUMBIA CHINOOK SALMON (ONCORHYNCHUS TSHAWYTSCHA) UNDER THE WILD SALMON POLICY

Chinook Salmon adult spawning phase. (Photo credit: Fisheries and Oceans Canada.)

Figure 1. Map of southern BC showing the Chinook Conservation Units.

Context: Canada’s Wild Salmon Policy’s (WSP) identifies six strategies for implementation. Strategy 1 is “Standardized monitoring of wild salmon status” and requires biological status assessments for all Pacific salmon conservation units (CUs). To conduct WSP status assessments, a toolkit comprised of a number of classes of indicators and metrics for status evaluation was completed in 2009. However, since a number of metrics can be used to evaluate biological status, it is possible that each metric can indicate a different status (Red, Amber, or Green). Therefore, status integration, which includes synthesis of CU status information across metrics into one or more status zones, and the provision of expert commentaries on the information used to assess status, is a useful final step in the status designation process. This report presents the application of WSP status integration conducted in a CSAS workshop. This workshop builds upon a previous application of WSP status integration techniques conducted for Fraser Sockeye CUs. This Science Advisory Report is from the February 4-6, 2014 Assessment of Southern British Columbia Chinook Salmon Conservation Units, Benchmarks and Status. Additional publications from this meeting will be posted on the Fisheries and Oceans Canada (DFO) Science Advisory Schedule as they become available.

August 2016 174 Pacific Region Biological Status of Southern BC Chinook Salmon

SUMMARY • A workshop entitled “Assessment of Southern British Columbia Chinook Salmon Conservation Units, Benchmarks and Status” was conducted to determine an integrated Wild Salmon Policy (WSP) status for each of the 35 southern BC Chinook Salmon Conservation Units (CU). The status integration method used was similar to that applied to Fraser Sockeye (Grant & Pestal 2013). A characteristic of southern BC Chinook Salmon CUs that is distinct from the Sockeye Salmon CUs assessed so far is the significant presence of hatchery-origin fish in addition to wild-origin fish in many of the CU area/watersheds. • For this workshop, multi-page standardized data summaries were produced for each southern BC Chinook Salmon CU. The data used to generate these summaries had been previously reviewed through two Regional Peer Review processes. • Participants were asked to determine a single WSP status zone from Red (poor status) to Amber (cautious status) to Green (healthy status) for the CU based on a combination of the information from the individual status metrics. • Status evaluations were completed and consensus reached on an integrated WSP status designation for 15 of the 35 CUs. Of these, 11 were assigned a Red status, one was assigned a Red/Amber status, one was assigned an Amber status and two were assigned a Green status. For another nine of the 35 CUs, an integrated status evaluation was not possible based on the information presented at the workshop. For these CUs, the status designation is “data deficient” and this designation is not expected to change until more information becomes available. For the remaining 11 of the 35 CUs, status evaluations were not completed. Instead, the status of these CUs was classified as “to be determined”. These CUs are a component of units where the enhanced sites are predominant; consensus was not reached on how to derive a WSP status assessment for such units. • In addition to providing final integrated status for each CU, the expert interpretation of the data summaries was documented in status commentaries. These commentaries provide the details underlying the final integrated status decisions. Status zones on their own do not provide an indication of which factors drive their designation, which would influence subsequent WSP strategies. The commentaries are an important source of information to inform management considerations. • The designation of seven Fraser River CUs as Red and two others with a status of Amber is especially noteworthy. A review of all Chinook populations in BC carried out more than 30 years ago found compelling evidence of substantial declines in abundance in all geographic regions, except within the Fraser River watershed. The last 12 to 15 years have been a period during which most groups of Chinook within the Fraser River have declined in numbers, and the outlook for Chinook outside of the Fraser River has generally not shown sustained improvement since the earlier review. • Integrated WSP status designations could not be developed for 20 of the 35 southern BC Chinook CUs based on the information and methods available to the workshop participants, which is very concerning. This highlights the need for additional work relating to information collection and assembly and for the development of a suitable method for status assessment when there is a significant contribution to recruitment and spawner abundance from enhanced sites. • A proposal on the frequency of status re-assessments was agreed to: DFO staff should recalculate the individual status metrics annually, update the standardized data summaries,

2 175 Pacific Region Biological Status of Southern BC Chinook Salmon

and check for any substantial changes. If results from individual metrics indicate a change that could affect the overall status for the CU, a meeting would be convened to address the affected CUs only. A full re-assessment of all CUs would take place every four years. INTRODUCTION The goal of the Wild Salmon Policy (WSP) is to “restore and maintain healthy salmon populations and their habitats for the benefit and enjoyment of the people of Canada in perpetuity” (DFO 2005). In order to achieve this goal, the WSP outlines a number of strategies, including Strategy 1 (Standardized Monitoring of Wild Salmon Status), which is the subject of this Science Advisory Report (SAR). Action Steps for Strategy 1 include: 1. identification of CUs; 2. development of criteria to assess CUs and identification of benchmarks to represent biological status; and, 3. monitoring and assessment of CU status. Work on these action steps has progressed since the WSP was published in 2005, with the following peer-reviewed milestones: • method for the identification of Pacific salmon CUs (Holtby & Ciruna 2007); • method for the assessment of Pacific salmon biological status under the WSP (Holt et al. 2009); • technical background for WSP status assessments (Holt 2009; Porszt 2009; Holt 2010; Holt & Bradford 2011; Porszt et al. 2012); • integration techniques for WSP status assessments of salmon CUs (Grant & Pestal 2013); • revision of southern BC Chinook Conservation Unit assignments (DFO 2013). Four classes of indicators have been recommended to evaluate WSP status of wild Pacific salmon: abundance, trends in abundance, distribution, and fishing mortality (Holt et al. 2009). Within each class of indicator, one or more metrics can be used for status assessments, and, for each metric, a lower benchmark and upper benchmark delineate the Red to Amber and Amber to Green status zones, respectively (Table 1). These biological benchmarks are specifically used for status assessments, and are not prescriptive for specific management actions. They are also designed to be more conservative than the criteria established by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC), as required by the WSP.

Table 1. The three zones of biological status defined in the WSP (Grant & Pestal 2013).

Status Definition

“… established at a level of abundance high enough to ensure there is a substantial buffer Red between it and any level of abundance that could lead to a CU being considered at risk of extinction by COSEWIC”

“While a CU in the Amber zone should be at low risk of loss, there will be a degree of lost Amber production. Still, this situation may result when CUs share risk factors with other, more productive units”

“identif[ies] whether harvests are greater than the level expected to provide on an average Green annual basis, the maximum annual catch for a CU, given existing conditions…there would not be a high probability of losing the CU”

3 176 Pacific Region Biological Status of Southern BC Chinook Salmon

Since CU status evaluations can include more than one metric, it is possible that different metrics could each indicate a different WSP status zone from Red (poor status) to Green (healthy status). For example, the WSP recent trend in abundance metric could suggest a CU’s status is poor, while conversely, the long-term trend metric could indicate the same CU’s status is healthy. In cases where metric information is contradictory, provision of this metric-specific status information alone does not provide complete scientific advice to fisheries management. Instead, a final step that synthesizes all metric and status-related information into an integrated status for each CU, and provides expert commentary on this information, is necessary as inputs into subsequent implementation of WSP Strategy 4 (Integrated Strategic Planning) to prioritize assessment activities and management actions (Table 2. Guidance in the WSP on assessment actions and management considerations for CUs in each of three status zones (Grant & Pestal 2013).Table 2).

Table 2. Guidance in the WSP on assessment actions and management considerations for CUs in each of three status zones (Grant & Pestal 2013).

Status Assessment Actions Management Considerations

“… a detailed analytical assessment will normally “Biological considerations will be the be triggered to examine impacts on the CU of primary driver for the management of CUs fishing, habitat degradation, and other human with Red status”. “The presence of a CU in factors, and evaluate restoration potential”, “… the Red zone will initiate immediate detailed stock assessments will identify the consideration of ways to protect the fish, Red reasons for the change in status”. “CUs in the increase their abundance, and reduce the Red zone … will be identified as management potential risk of loss”. priorities … the protection and restoration of these CUs will be primary drivers for harvest, habitat, and enhancement planning.”

“… a detailed analytical assessment may be “Decisions about the conservation of CUs required to input into Strategies 2 & 3..” in the Amber zone will involve broader considerations of biological, social, and economic issues”; “involves a comparison Amber of the benefits from restoring production versus the costs arising from limitations imposed on the use of other CUs to achieve that restoration”; “implies caution in the management of the CU”

“ a detailed analytical assessment of its biological “Social and economic considerations will status will not usually be needed” tend to be the primary drivers for the management of CUs in the green zone, Green though ecosystem or other non- consumptive values could also be considered”.

For Pacific Salmon CUs, WSP biological status integration methods have previously been developed and applied to Sockeye Salmon assessments (Grant & Pestal 2013). However, a characteristic of southern BC Chinook Salmon CUs that is distinct from the Sockeye Salmon CUs assessed so far is that many areas support substantial numbers of hatchery-origin fish in addition to wild-origin fish. Therefore, the guidelines developed for Sockeye Salmon are only partially applicable to the southern BC Chinook Salmon situation. In order to explore the applicability of the status integration techniques developed previously, and to provide WSP

4 177 Pacific Region Biological Status of Southern BC Chinook Salmon status assessments, a CSAS workshop entitled “Assessment of Southern British Columbia Chinook Salmon Conservation Units, Benchmarks and Status” was conducted to achieve these goals. This SAR summarizes the results from this CSAS workshop. The objectives of the workshop were to: 1. Determine an integrated WSP status for each southern BC Chinook Salmon CU; 2. Indicate the effect on the status assessments of including, or excluding, enhanced Chinook Salmon contributions; 3. Provide advice on data and methods required for assessing the status of any CUs that are currently data deficient; 4. Include information specific to each CU on fishing mortality, where possible; 5. Provide advice on the appropriate frequency of status re-assessment, changes to monitoring variables that could invoke early re-assessment, and the appropriate timing for assessment relative to data availability; and 6. Identify and recommend data management approaches required to support recommended changes to re-assessment of CUs. ASSESSMENT

Data For this workshop, multi-page standardized data summaries were produced for each southern BC Chinook Salmon CU. The data used to generate these summaries had been previously reviewed through two CSAS Regional Peer Review processes1,2. These data summaries included the following: • time series plots of spawner abundances (either relative indices or absolute abundances, where available); • a table of absolute abundances relative to COSEWIC criteria D1 for small populations; • a summary of overall data quality (as a percentage of spawner abundance); • a summary of the categorization of enhancement activity level by census site3;

1Brown, G.S., Baillie, S.J., Thiess, M.E., Bailey, R.E., Candy, J.R., Parken, C.K., and Willis, D.M. 2014. Pre-COSEWIC Review of Southern British Columbia Chinook Salmon (Oncorhynchus tshawytscha) Conservation Units: Part I, Background. CSAS Working Paper 2012/P62. In revision.

2 Brown, G.S., Baillie, S.J., Bailey, R.E., Candy, J.R., Holt, C.A, Parken, C.K., Pestal, G.P., Thiess, M.E., and Willis, D.M. 2014. Pre-COSEWIC Review of Southern British Columbia Chinook Salmon (Oncorhynchus tshawytscha) Conservation Units, Part II: Data, Analysis and Synthesis. CSAS Working Paper 2012/13 P23. In revision.

3 The concepts of a “Total Unit” (TU) and an Enhancement Unit (EU) were introduced at the workshop. A Total Unit can be comprised of two components: the CU and an associated EU. The CU includes only census sites with low or unknown enhancement level activity in an attempt to be consistent with the WSP focus on ‘wild salmon’. The EU contains only census sites with moderate or high enhancement level activity. Although these concepts were introduced at the workshop, they were not endorsed by the participants and therefore are not considered to form a viable conceptual approach to this issue.

5 178 Pacific Region Biological Status of Southern BC Chinook Salmon

• a stacked bar plot illustrating the distribution of spawner abundance across sites within the CU; • a whisker plot illustrating short term trends by census site within the CU; • a table of spawner abundance by census site within the CU; • status information for up to three WSP metrics: one metric for abundance relative to biological benchmarks, one metric for extent of decline in abundance, and two related metrics for short-term trend in abundance; • where available, supplementary time series plots of natural log-transformed spawner abundance, generational average of spawner abundance, CWT indicator spawner abundance, total return, productivity (recruits/spawner by brood year), hatchery releases from within and outside the CU, exploitation rates and marine survival; • retrospective (historical) time series of status for each WSP metric relative to established benchmarks (Holt et al. 2009).

Methods Workshop participants were invited to attend based on their experience with different aspects of salmon assessment and included DFO staff from Science, Ecosystems Management and Fisheries Management sectors and external participants from First Nations organizations, the commercial and recreational fishing sectors, environmental non-governmental organizations, and academia. Participants were requested to join one of four pre-workshop seminars in order to review the data summary layout and to provide feedback to organizers on the workshop format. At the workshop, participants were assigned to one of six groups, each comprised of six or seven individuals. Their group assignment remained the same for the duration of the workshop. Individuals were assigned in order to provide a varied mix of views and expertise within each group. Each of the 35 CUs (and their associated enhanced sites where applicable) was designated as an individual case study. The identity of the CU represented by a case study was not revealed to the participants during the initial assessment sessions. This “blind” approach was similar to that employed by Grant & Pestal (2013) during the Fraser Sockeye workshop. The 35 case studies were presented in seven sets over the first two days of the workshop. Participant groups were given 15 minutes, 30 minutes, one hour or 1.5 hours, depending on the set size and complexity, to discuss each set in a breakout session. At the end of each breakout session, back in a full participant plenary session, groups compared results and discussed their reasoning for their final integrated status designations. All of the 35 CUs were evaluated by at least some of the groups, and each group evaluated a representative number of CU types (different metrics and statuses). Late on the second day, the CU identity of each case study was revealed to the participants. The third day of the workshop was a full day of plenary discussion to reconcile group integrated status results allowing for use of knowledge of the identity of each CU.

Results Final Integrated Status By the end of the workshop, participants completed status evaluations and reached consensus on an integrated WSP status designation for 15 of the 35 CUs (Table 3 and Figure 2). The 15 southern BC Chinook CUs are ordered in Table 3 using their final integrated status, with CUs designated Red (poorest status) located at the top of the table to CUs designated Green (best

6 179 Pacific Region Biological Status of Southern BC Chinook Salmon status) at the bottom. Thirteen out of the 15 CUs were reconciled between groups in the post- reveal plenary session to a single WSP status zone. There was one CU where final integrated statuses included two status zones. The Lower Fraser River_FA_0.3 (CK-03) CU’s integrated Green status was flagged as provisional by participants. Following the example of the Fraser Sockeye WSP status assessments (Grant & Pestal 2013), when some participants held divergent views, the status assignment was classified as “provisional”. In this case, the short- term decline observed in recent years, despite decreasing exploitation rate, resulted in a provisional status designation to highlight the need for monitoring the trend. For another nine of the 35 CUs, an integrated status evaluation was not possible based on the information presented at the workshop. For these CUs, the status designation is “data deficient” (DD). When preparing the data summaries, the workshop organizers identified five CUs as obviously data deficient (Table 3, Cases 31 to 35). The workshop participants supported this initial assessment and also designated an additional four CUs as data deficient. For all nine of these CUs, the status designation is not expected to change until more information is available. South Thompson-Bessette Creek_SU_1.2 (CK-16) and Okanagan_1.x (CK-01) were designated as Red status. However, there was some concern expressed by the participants that the definition of these CUs might not be valid. The status of these CUs should be re-evaluated following a review of their CU definitions. The remaining 11 of the 35 CUs (Table 4) presented a substantial challenge for the participants and ultimately, status evaluations could not be completed for them. Instead, the status of these CUs was classified as “to be determined” (TBD). These CUs are geographically proximate to predominantly enhanced sites, or data exist only for the enhanced sites geographically proximate to the CU (e.g. a CU may exist but no wild census sites have data of sufficient quality for assessment at this time). Consensus was not reached on how to derive a WSP status assessment for such combined wild and enhanced site units, or the CUs that spawn in the same area. A method to consider enhanced contribution by redefining the wild site versus enhanced site classification in the data summaries was proposed by the workshop organizers. However; there was consensus that a review of the proposed method was not within the scope of the workshop and should be the subject of a future review. Although there are no status evaluations provided for these 11 CUs, unlike the situation with the data deficient CUs, an integrated WSP status could be determined in some cases once a suitable method is developed to assess the status of enhanced sites and how they should be considered in status assessments of the CU. Status Commentaries In addition to documenting a final integrated status designation for each CU, the expert interpretation of the data summaries was recorded as status commentaries (Appendix B of the Research Document resulting from the workshop). These commentaries provide the details underlying the final integrated status decisions, which varied even amongst CUs with identical status designations. These details will be important when the results from Strategy 1 (Standardized Monitoring of Wild Salmon Status) are linked to Strategy 4 (Integrated Strategic Planning). Status zones on their own do not provide an indication of which factors drive their designation, which would influence subsequent WSP strategies. The commentaries are an important source of information to inform management considerations.

7 180 Pacific Region Biological Status of Southern BC Chinook Salmon

Table 3. Summary of completed integrated status evaluations for Southern BC Chinook Salmon CUs.

Integrated status evaluation completed at workshop Integrated Case Status # CU ID CU Name Area RED 1 CK-10 Middle Fraser River_SP_1.3 Fraser RED 4 CK-18 North Thompson_SP_1.3 Fraser RED 6 CK-19 North Thompson_SU_1.3 Fraser RED 11 CK-09 Middle Fraser River-Portage_FA_1.3 Fraser RED 24 CK-17 Lower Thompson_SP_1.2 Fraser RED 25 CK-31 West Vancouver Island-South_FA_0.x WCVI RED 26 CK-12 Upper Fraser River_SP_1.3 Fraser RED 29 CK-29 East Vancouver Island-North_FA_0.x Inner SC RED 30 CK-32 West Vancouver Island-Nootka & Kyuquot_FA_0.x WCVI RED* 3 CK-16 South Thompson-Bessette Creek_SU_1.2 Fraser RED* 5 CK-01 Okanagan_1.x Columbia RED / AMBER 27 CK-14 South Thompson_SU_1.3 Fraser AMBER 12 CK-11 Middle Fraser River_SU_1.3 Fraser GREEN(p) 9 CK-03 Lower Fraser River_FA_0.3 Fraser GREEN 2 CK-13 South Thompson_SU_0.3 Fraser

Integrated status evaluation not possible based on information presented at workshop Integrated Case Status # CU ID CU Name Area DD 7 CK-82 Upper Adams River_SU_x.x Fraser DD 8 CK-06 Lower Fraser River_SU_1.3 Fraser DD 10 CK-05 Lower Fraser River-Upper Pitt_SU_1.3 Fraser DD 28 CK-28 Southern Mainland-Southern Fjords_FA_0.x Inner SC DD 31 CK-08 Middle Fraser-Fraser Canyon_SP_1.3 Fraser DD 32 CK-20 Southern Mainland-Georgia Strait_FA_0.x Inner SC DD 33 CK-34 Homathko_SU_x.x Inner SC DD 34 CK-23 East Vancouver Island-Nanaimo_SP_1.x Inner SC DD 35 CK-35 Klinaklini_SU_1.3 Inner SC “(p)” means provisional, and identifies cases where some participants held divergent views. “*” means that CU definition should be reviewed.

8 181 Pacific Region Biological Status of Southern BC Chinook Salmon

Figure 2. Map of southern BC summarizing workshop consensus on biological status of southern BC Chinook Salmon CUs.

Status Integration Approaches The workshop organizers had prepared an initial set of guidelines for status integration (see Appendix E of the Proceedings resulting from the workshop). These guidelines were largely based on the recommendations in Grant and Pestal (2013). After the groups had completed several evaluations they reported that they were adopting patterns in their approach to status integration. Based on the feedback from participants, the guidelines were revised and are reported in Section 3 of the Research Document resulting from the workshop. In addition, the status deliberation notes and plenary discussions exposed some common themes to status integration approaches that were not explicitly endorsed as guidelines by the participants. These are also documented in Section 3 of the Research Document resulting from the workshop.

9 182 Pacific Region Biological Status of Southern BC Chinook Salmon

Table 4. Summary of incomplete integrated status evaluations for Southern BC Chinook Salmon CUs. Integrated status evaluation not attempted at workshop due to unresolved methods Integrated Case Status # CU ID CU Name Area TBD** 13 CK-04 Lower Fraser River_SP_1.3 Fraser TBD 14 CK-21 East Vancouver Island-Goldstream_FA_0.x Inner SC TBD 15 CK-33 West Vancouver Island-North_FA_0.x WCVI TBD 16 CK-22 East Vancouver Island-Cowichan & Koksilah_FA_0.x Inner SC TBD 17 CK-02 Boundary Bay_FA_0.3 Inner SC TBD 18 CK-07 Maria Slough_SU_0.3 Fraser TBD 19 CK-25 East Vancouver Island-Nanaimo & Chemainus_FA_0.x Inner SC TBD 20 CK-15 Shuswap River_SU_0.3 Fraser TBD 21 CK-83 East Vancouver Island-Georgia Strait_SU_0.3 Inner SC TBD 22 CK-27 East Vancouver Island-Qualicum & Puntledge_FA_0.x Inner SC TBD 23 CK-9008 Fraser-Harrison fall transplant_FA_0.3 Fraser “**” means that CU status should be re-evaluated after review of enhancement level definition.

Sources of Uncertainty • The standardized data summaries were prepared based on data that had been previously reviewed1,2, however, these summaries are based largely on spawner data with a substantial but unquantified level of uncertainty. • A period of apparent abundance increases occurred during the 1990s and early 2000s when major improvements were made in many BC escapement programs. These escapement estimation improvements typically resulted in immediate and noticeably higher annual estimates relative to earlier estimates. This suggests that apparent improvements in abundance could be related more to changes in survey and estimation methods than to genuine biological changes. • Some of the abundance time series represent relative rather than absolute abundances. Relative abundances likely under-estimate true abundance (by unknown and variable amounts), so an indication of red zone status in relation to the WSP metric on absolute abundance may not be accurate. • Some of the individual metrics display a pattern of changing status from one year to the next (e.g. red status one year followed by green status the next year and then returning to red). In this situation, the metric is not conveying meaningful results for determining integrated status and would typically be disregarded or given less weight in status deliberations. • Information on the contribution of enhanced fish to the abundance of fish observed at “wild” sites is often limited; and as such, the actual wild contribution (which is key to the WSP CU definition) is often unknown. For the purposes of these status assessments, observations at wild sites are assumed to be comprised entirely of wild fish. • The status evaluations developed at this workshop ultimately relied on the expert opinions of the participants and as such, are subject to the experience and opinions of the individuals involved. Because many of the evaluations are more subjective than objective, the

10 183 Pacific Region Biological Status of Southern BC Chinook Salmon

repeatability of these findings is uncertain. The status commentaries in Appendix B of the Research Document resulting from the workshop identify cases where participants were especially confident in their assessment, as well as cases where the status designations were particularly uncertain, which may be useful in developing approaches to quantifying this uncertainty in the future. CONCLUSIONS AND ADVICE

Southern BC Chinook CUs Integrated Status Integrated status designations were developed for 15 of the 35 southern BC Chinook CUs, and status commentaries were provided for all 35 CUs. In some cases, the commentaries provide more useful advice for management considerations than would be indicated by the mapping of the status zone to the management considerations in Table 2. These results address two of the six objectives for the workshop: “determine an integrated WSP status for each southern BC Chinook Salmon CU”, and “include information specific to each CU on fishing mortality, where possible”. The majority of CUs for which an integrated status was developed occurred within the Fraser River watershed (11 of 15). This reflects the reduced prevalence of enhancement as a management intervention in that region. While seven of the Fraser River CUs were designated as Red, all four of the CUs that were assessed from other regions were also designated as Red. All adult and juvenile life history patterns known in southern BC Chinook are represented in the group of 11 Red status CUs. This suggests that declines in abundance shown by these CUs cover a broad geographic area and are not specific to any particular group of Chinook Salmon. The designation of seven Fraser River CUs as Red and two others with a status of Amber is especially noteworthy. A review of all Chinook populations in BC carried out by Healey (1982) more than 30 years ago found compelling evidence of substantial declines in abundance in all geographic regions, except within the Fraser River watershed. Riddell et al. (2013) suggested that spawner abundances in most southern BC areas may have increased for a period in the 1990s and early 2000s. However, these apparent improvements in abundance could be related more to changes in survey and estimation methods than to genuine biological changes. Regardless of whether real abundance increases occurred in the 1990s, the last 12 to 15 years have been a period during which most groups of Chinook within the Fraser River have declined in numbers. The outlook for Chinook Salmon outside of the Fraser River has generally not shown sustained improvement since Healey’s (1982) review. Recommendations • Integrated status designations could not be developed for 20 of the 35 southern BC Chinook CUs based on the information and methods available to the workshop participants. This represents the majority of the southern BC Chinook CUs, or approximately 21% of the surveyed aggregate abundance, which is a concern. This highlights the need for additional work and relates to the objectives: “provide advice on data and methods required for assessing the status of any Conservation Units that are currently data deficient”, and “identify and recommend data management approaches required to support recommended changes to re-assessment of CUs”. • In some cases, additional information relating to the data deficient CUs is in the possession of the Department, but has not yet been incorporated into the regional escapement data holdings where it would be accessible to analysts. If this information were incorporated, it is possible that some of the CUs would no longer be data deficient and status designations could be developed. This information includes escapement survey records held by local

11 184 Pacific Region Biological Status of Southern BC Chinook Salmon

offices in paper and electronic formats that have not been a priority for further analysis to date. The work necessary to locate and incorporate this information into the regional escapement data holdings could provide significant benefits for future status assessments. • The workshop participants identified an issue where a Chinook population is known anecdotally to exist, but there are no escapement surveys recorded in the regional escapement data holdings. Examples of this are information from local traditional knowledge, data from non-DFO programs such as fish habitat surveys initiated for forestry purposes, and data from juvenile salmon surveys. Since the regional adult escapement data holdings provided the source information for initial CU definition, the absence of survey records meant that these populations were not included in the CU definitions. Thus it is possible that there are additional Chinook CUs yet to be defined. These would likely form additional CUs for the data deficient category. This issue could be addressed by incorporating the information on un-surveyed but known Chinook populations into the regional escapement data holdings as placeholder records. • The amount of data filtered out due to data quality concerns prior to status assessments raises questions regarding the utility of temporally extensive, low-quality surveys and their role in the stock assessment program should be reviewed. If such data are not useful for status assessment, then they are of little value other than indicating fish presence which has proved useful only in identifying spawning sites for potential grouping within a CU. • Aside from the data deficiency issue, the other issue which prevented integrated status designations relates to the workshop objective: “indicate the effect on the status assessments of including and excluding enhanced Chinook Salmon, where applicable”. This was the only objective of the workshop that was not successfully addressed. The participants attempted to address this objective but the consensus was that given the methods and guidelines available to them, status designation was not possible for CUs that had a substantial contribution from enhanced sites. To resolve this issue for future assessments would require a specific project to develop a suitable method for status assessment for sites (or groups of sites) with significant enhancement contribution. In addition, guidance would need to be developed for considering the interaction between the CU and an associated enhanced contribution in the status assessment of the CU. The resulting proposed method and guidelines should then be subject to peer review. Once this work is complete, the southern BC Chinook CUs currently categorized with a status of To Be Determined should be re-assessed.

Status Integration Process Again, similar to the approach taken for Fraser Sockeye Salmon CUs (Grant & Pestal 2013), expert opinion on status integration and associated commentaries were elicited through a combination of smaller breakout groups and full participant plenary sessions. The advantage of this approach was that it permitted independent small-group evaluation of a range of integration approaches and integrated status designations, which could then be consolidated in a plenary session with all participants. Although not highlighted in the results presented here, more often than not, the individual group results showed a similar status designation for a CU and the status reconciliation during the plenary session was rapid and not controversial. This provides some confidence that the integration process is more objective than subjective, and is repeatable.

12 185 Pacific Region Biological Status of Southern BC Chinook Salmon

Integration Guidelines Now that two of these larger integration workshops have occurred, and a variety of CUs have been examined, it might be possible to prepare a more comprehensive set of integration guidelines for formal peer-review. Once accepted, these guidelines could allow for the completion of a preliminary status integration report for a collection of CUs by a small expert team. This report would then become the working paper to be reviewed via the more typical CSAS Regional Peer Review process. If this work were undertaken it would help to address the concern that the workshop format for WSP status assessment is onerous and is limiting the opportunity for status assessments.

Frequency of Re-Assessment A key workshop objective was to “provide advice on the appropriate frequency of status re- assessment, changes in monitoring variables that could invoke early re-assessment, and appropriate timing for assessment relative to data availability”. The following proposal on the frequency of status re-assessments was agreed on by participants in plenary session. • DFO staff should recalculate the individual status metrics annually, update the standardized data summaries, and check for any substantial changes. • A meeting would not be required to re-assess status of CUs unless results from individual metrics indicated a change that could affect the overall status for the CU. • A shorter (and perhaps smaller) meeting would be convened to address the affected CUs only. • A full re-assessment of all CUs would take place every four years (representing approximately once per generation for most Chinook CUs). • Full re-assessment meetings would include representation from DFO and stakeholders, but could be shorter than the current workshop; the meeting could review a status assessment working paper, and could possibly be vetted through a CSAS Science Response process instead of a Regional Peer Review process.

SOURCES OF INFORMATION This Science Advisory Report is from the February 4-6, 2014 Assessment of Southern British Columbia Chinook Salmon Conservation Units, Benchmarks and Status. Additional publications from this meeting will be posted on the Fisheries and Oceans Canada (DFO) Science Advisory Schedule as they become available. DFO. 2005. Canada's Policy for Conservation of Wild Pacific Salmon. Fisheries and Oceans Canada, Vancouver, BC. 34 pp. (Accessed 12 January 2016) DFO. 2013. Review and Update of Southern BC Chinook Conservation Unit Assignments. DFO Can. Sci. Advis. Sec. Sci. Resp. 2013/022. 25 pp. (Accessed 06 July 2016) Grant, S.C.H. & G. Pestal. 2013. Integrated Biological Status Assessments Under the Wild Salmon Policy Using Standardized Metrics and Expert Judgement: Fraser River Sockeye Salmon (Oncorhynchus nerka) Case Studies. Can. Sci. Advis. Sec. Res. Doc. 2012/106. v + 132 p. (Accessed 06 July 2016) Healey, M.C. 1982. Catch, Escapement and Stock-Recruitment for British Columbia Chinook Salmon Since 1951. Can. Tech. Rep. Fish. Aquat. Sci. 1107: iv + 77 p. (Accessed 06 July 2016)

13 186 Pacific Region Biological Status of Southern BC Chinook Salmon

Holt, C.A. 2009. Evaluation of Benchmarks for Conservation Units in Canada's Wild Salmon Policy: Technical Documentation. Can. Sci. Advis. Sec. Res. Doc. 2009/059. xii + 50 pp. (Accessed 06 July 2016) Holt, C.A. and Bradford, M.J. 2011. Evaluating Benchmarks of Population Status for Pacific Salmon. N. Am. J. Fish. Manage. 31:2, 363-378. (Accessed 06 July 2016) Holt, C.A., Cass, A., Holtby, B., & Riddell, B. 2009. Indicators of Status and Benchmarks for Conservation Units in Canada's Wild Salmon Policy. Can. Sci. Advis. Sec. Res. Doc. 2009/058. vii + 74 pp. (Accessed 12 January 2016) Holt, C.A. 2010. Will Depleted Populations of Pacific Salmon Recover Under Persistent Reductions in Survival and Catastrophic Mortality Events? ICES J. Mar. Sci. 67 (9): 2018- 2026. (Accessed 06 July 2016) Holtby, L.B. & Ciruna, K.A. 2007. Conservation Units for Pacific Salmon Under the Wild Salmon Policy. Can. Sci. Advis. Sec. Res. Doc. 2007/070. viii + 350 pp. (Accessed 06 July 2016) Porszt, E.J. 2009. An Evaluation of Criteria for Assessing Conservation Status of Fraser Sockeye Conservation Units. M.R.M. thesis, School of Resource and Environmental Management, Simon Fraser University, Burnaby, B.C. xiii + 104 pp. (Accessed 06 July 2016) Porszt, E.J., Peterman, R.M., Dulvy, N.K., Cooper, A.B. and Irvine, J.R. 2012. Reliability of Indicators of Decline in Abundance. Conservation Biology. Doi:1111/j.1523- 1739.2012.01882.x (Accessed 06 July 2016) Riddell, B., M. Bradford, R. Carmichael, D. Hankin, R. Peterman, and A. Wertheimer. 2013. Assessment of Status and Factors for Decline of Southern BC Chinook Salmon: Independent Panel’s Report. Prepared with the assistance of D.R. Marmorek and A.W. Hall, ESSA Technologies Ltd., Vancouver, B.C. for Fisheries and Oceans Canada (Vancouver. BC) and Fraser River Aboriginal Fisheries Secretariat (Merritt, BC). xxix + 165 p. + Appendices. (Accessed 06 July 2016)

14 187 Pacific Region Biological Status of Southern BC Chinook Salmon

THIS REPORT IS AVAILABLE FROM THE: Center for Science Advice Pacific Region Fisheries and Oceans Canada 3190 Hammond Bay Road Nanaimo, BC V9T 6N7 Telephone: 250-756-7208 E-Mail: [email protected] Internet address: www.dfo-mpo.gc.ca/csas-sccs/ ISSN 1919-5087 © Her Majesty the Queen in Right of Canada, 2016

Correct Citation for this Publication: DFO. 2016. Integrated Biological Status of Southern British Columbia Chinook Salmon (Oncorhynchus tshawytscha) Under the Wild Salmon Policy. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2016/042. Aussi disponible en français : MPO. 2016. État biologique intégré du saumon quinnat (oncorhynchus tshawytscha) du sud de la Colombie-Britannique en vertu de la politique concernant le saumon sauvage. Secr. can. de consult. sci. du MPO, Avis sci. 2016/042.

15 188 Species at Risk Act Recovery Strategy Series

Recovery Strategy for the Northern and Southern Resident Killer Whales (Orcinus orca) in Canada

Killer Whale

Photo: G. Ellis

Original publication 2008 1st amendment 2011 2nd amendment 2018

189 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Recommended citation:

Fisheries and Oceans Canada. 2018. Recovery Strategy for the Northern and Southern Resident Killer Whales (Orcinus orca) in Canada. Species at Risk Act Recovery Strategy Series, Fisheries & Oceans Canada, Ottawa, x + 84 pp.

For copies of the recovery strategy, or for additional information on species at risk, including Committee on the Status of Endangered Wildlife in Canada (COSEWIC) Status Reports, residence descriptions, action plans, and other related recovery documents, please visit the SAR Public Registry.

Cover illustration: Graeme Ellis, Fisheries & Oceans Canada

Également disponible en français sous le titre : « Programme de rétablissement des épaulards résidents (Orcinus orca) du nord et du sud au Canada»

Content (excluding the cover illustration) may be used without permission, with appropriate credit to the source.

ii 190 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Preface

The federal, provincial, and territorial government signatories under the Accord for the Protection of Species at Risk (1996) agreed to establish complementary legislation and programs that provide for effective protection of species at risk throughout Canada. Under the Species at Risk Act (S.C. 2002, c.29) (SARA), the federal competent ministers are responsible for the preparation of a recovery strategy for species listed as extirpated, endangered, or threatened and are required to report on progress five years after the publication of the final document on the Species at Risk Public Registry.

The Minister of Fisheries and Oceans and the Minister responsible for the Parks Canada Agency are the competent ministers under SARA for the Northern and Southern Resident Killer Whale and have prepared this strategy, as per section 37 of SARA. In preparing this recovery strategy, the competent ministers have considered, as per section 38 of SARA, the commitment of the Government of Canada to conserving biological diversity and to the principle that, if there are threats of serious or irreversible damage to the listed species, cost-effective measures to prevent the reduction or loss of the species should not be postponed for a lack of full scientific certainty. To the extent possible, the recovery strategy has been prepared in cooperation with Environment and Climate Change Canada and the Province of British Columbia, as per section 39(1) of SARA. In addition, both populations are considered trans-boundary in United States waters. The U.S. National Oceanic and Atmospheric Administration also participated in its preparation.

As stated in the preamble to SARA, success in the recovery of this species depends on the commitment and cooperation of many different constituencies that will be involved in implementing the directions set out in this strategy and will not be achieved by Fisheries and Oceans Canada, the Parks Canada Agency, or any other jurisdiction alone. The cost of conserving species at risk is shared amongst different constituencies. All Canadians are invited to join in supporting and implementing this strategy for the benefit of the Northern and Southern Resident Killer Whale and Canadian society as a whole.

In addition to this recovery strategy, an action plan for the Northern and Southern Resident Killer Whales has been developed that provides information on recovery measures underway and to be taken by Fisheries and Oceans Canada, the Parks Canada Agency and other jurisdictions and organizations involved in the conservation of the species. Implementation of this strategy is subject to appropriations, priorities, and budgetary constraints of the participating jurisdictions and organizations. Multi-species action plans have also been developed by the Parks Canada Agency that include recovery measures for Northern and Southern Resident Killer Whales.

Acknowledgements

Fisheries and Oceans Canada wishes to thank the Resident Killer Whale Recovery Team (Appendix B) for its efforts in developing the 2008 recovery strategy for the Northern and Southern Resident Killer Whales in Canada. The Team members were generous in contributing their own time to the development of the proposed recovery strategy. The recovery strategy is based on an extensive literature review and on technical input from individual team members

iii 191 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

and from group discussions and was mostly written by Kathy Heise. The Recovery Team is grateful for the expert reviews provided by Dr. Volker Deecke of the University of British Columbia and Dr. Christophe Guinet, Centre d’Etudes Biologiques de Chize, France. The cover photo was provided by Graeme Ellis. Doug Sandilands (Vancouver Aquarium Marine Science Centre) provided Figure 1. The amendments in the 2011 recovery strategy were prepared by Fisheries and Oceans Canada. The 2018 amendment to the recovery strategy was prepared by Christie McMillan with input from Dr. John Ford, Dr. Sheila Thornton, Dr. Thomas Doniol- Valcroze, Dr. Lisa Jones, Heather Brekke, Karen Leslie, and Dr. Lance Barrett-Lennard. Robin Abernethy prepared Figures 4 and 5.

iv 192 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Strategic environmental assessment statement

A strategic environmental assessment (SEA) is conducted on all SARA recovery planning documents, in accordance with the Cabinet Directive on the Environmental Assessment of Policy, Plan and Program Proposals. The purpose of a SEA is to incorporate environmental considerations into the development of public policies, plans, and program proposals to support environmentally-sound decision making.

Recovery planning is intended to benefit species at risk and biodiversity in general. However, it is recognized that strategies may also inadvertently lead to environmental effects beyond the intended benefits. The planning process based on national guidelines directly incorporates consideration of all environmental effects, with a particular focus on possible impacts on non- target species or habitats. The results of the SEA are incorporated directly in the strategy itself, but are also summarized below.

While this recovery strategy will clearly benefit the environment by promoting the recovery of the Northern and Southern Resident Killer Whales, several potentially adverse effects also were considered. Through the development of this strategy numerous anthropogenic factors that jeopardize or have potential to jeopardize the recovery of these populations were evaluated and are presented. Principal among the anthropogenic factors or threats are environmental contamination, reductions in the availability or quality of prey, and both physical and acoustic disturbance. In some cases these factors threaten the populations; in other cases they may affect critical habitat. It was concluded that some threats can be mitigated through the use of existing legislation, policies, and programs and, in fact, there are numerous examples of mitigation measures that are currently employed outlined herein. However, in other cases the threat and/or the potential mitigation measure(s) require further research or evaluation before recommendations on specific actions or activities can be formulated. The general type of research, evaluation, and approaches for mitigation are presented in this strategy. However, through the course of action planning, specific activities for recovery and mitigation have been evaluated and detailed in the action plan for these populations along with an evaluation of effects and costs of these activities and measures. Therefore, taking into account the general nature of the recommendations for mitigation to recover these populations and that many of the recommendations to protect critical habitat fall under existing legislation and policies, this strategy will not entail any new significant adverse effects.

v 193 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Executive summary

Two distinct populations of Resident Killer Whales (Orcinus orca), known as the Northern and Southern Residents, occupy the waters off the west coast of Canada. In 2001, the Committee on the Status of Endangered Wildlife in Canada (COSEWIC) designated Southern Resident Killer Whales as Endangered and Northern Resident Killer Whales as Threatened. Both populations are listed in Schedule 1 of the Species at Risk Act (SARA). These two populations are acoustically, genetically, and culturally distinct.

The “Recovery Strategy for the Northern and Southern Resident Killer Whales (Orcinus orca) in Canada” was finalized and published on the Species at Risk Public Registry in 2008. Minor amendments to the recovery strategy were made in 2011 to provide additional clarification regarding critical habitat for Northern and Southern Resident Killer Whales. This recovery strategy is herewith amended once again to include identification of additional critical habitat for these populations and to provide minor updates to background and species information. Additionally, sections of the recovery strategy have been reordered to align with current templates. This recovery strategy is considered one in a series of documents for this species that are linked and should be taken into consideration together; including the COSEWIC status report (COSEWIC 2008) and the Action Plan for the Northern and Southern Resident Killer Whale (Orcinus orca) in Canada (DFO 2017a). Recovery has been determined to be biologically and technically feasible.

Resident Killer Whale populations in Canadian Pacific waters are presently considered to be at risk because of their small population size, low reproductive rate, and the existence of a variety of anthropogenic threats that have the potential to prevent recovery or to cause further declines. Principal among these anthropogenic threats are environmental contamination, reductions in the availability or quality of prey, and both physical and acoustic disturbance. Even under the most optimistic scenario (human activities do not increase mortality or decrease reproduction), the species’ low intrinsic growth rate means that the time frame for recovery will be more than one generation (25 years).

The Southern Resident Killer Whale population has fluctuated between 70 and 99 individuals since 1976, and consisted of 74 members in 2018 (Center for Whale Research unpublished data). During the summer and fall, Southern Residents are primarily found in the trans-boundary waters of Haro Strait, Boundary Pass, the eastern portion of Juan de Fuca Strait, and southern portions of the Strait of Georgia. Some members of the population typically remain in the same general area in winter and spring, but others range over much greater distances, and have been reported as far south as central California, and as far north as southeastern Alaska. During the summer and fall, the principal prey of Southern Residents is Chinook and Chum Salmon (Oncorhynchus tshawytscha and O. keta); less is known of their diet in the winter and spring.

The Northern Resident Killer Whale population experienced a decline of 7% between 1997 and 2002, but since that time has been increasing at a mean rate of 2.9% per year, reaching approximately 309 individuals in 2017 (Towers et al. 2015; Fisheries and Oceans Canada Cetacean Research Program (DFO-CRP) unpublished data). The population appears to spend the majority of its time from Campbell River and Alberni Inlet northwest to Dixon Entrance, but has been sighted as far south as Grays Harbor, Washington, and as far north as Glacier Bay, Alaska (Ford et al. 2000; 2017). Northern Residents also feed primarily on Chinook and Chum Salmon during the summer and fall. However, like Southern Residents, less is known of their

vi 194 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

winter distribution and diet, and this knowledge gap must be addressed to fully understand the principal threats affecting the population.

The goal of the Resident Killer Whale recovery strategy is to: “ensure the long-term viability of Resident Killer Whale populations by achieving and maintaining demographic conditions that preserve their reproductive potential, genetic variation, and cultural continuity1”.

In order to achieve this goal, four principal objectives have been identified. They are:

Objective 1: ensure that Resident Killer Whales have an adequate and accessible food supply to allow recovery

Objective 2: ensure that chemical and biological pollutants do not prevent the recovery of Resident Killer Whale populations

Objective 3: ensure that disturbance from human activities does not prevent the recovery of Resident Killer Whales

Objective 4: protect critical habitat for Resident Killer Whales and identify additional areas for critical habitat designation and protection

A description of the broad strategies to be taken to address threats to the species’ survival and recovery, as well as research and management approaches needed to meet the recovery goal and objectives are included in section 6. These strategies helped to inform the development of specific recovery measures in the action plan for Northern and Southern Resident Killer Whales. However, significant gaps in knowledge about Killer Whales remain and numerous actions have been identified to address these knowledge gaps and to identify further directions for recovery.

For Northern and Southern Resident Killer Whales, critical habitat is identified to the extent possible using the best available information, and provides the functions, features, and attributes necessary to support the species’ life-cycle processes and contribute to achieving the species’ recovery goal and objectives. This recovery strategy identifies critical habitat for Resident Killer Whales as four geographical areas: 1) the waters of Johnstone Strait and southeastern Queen Charlotte Strait (Northern Resident Killer Whale critical habitat); 2) transboundary waters in southern British Columbia, including southern Georgia Strait, Haro Strait, and Juan de Fuca Strait (Southern Resident Killer Whale critical habitat); 3) waters on the continental shelf off southwestern Vancouver Island, including Swiftsure and La Pérouse Banks (Northern and Southern Resident Killer Whale critical habitat); and 4) waters of west Dixon Entrance, along the north coast of Graham Island from Langara to Rose Spit (Northern Resident Killer Whale critical habitat) (section 7).

The Action Plan for the Northern and Southern Resident Killer Whale (Orcinus orca) in Canada was finalized and posted on the Species at Risk Public Registry in 2017. Additionally, multi- species action plans developed by the Parks Canada Agency include recovery measures for Northern and Southern Resident Killer Whales.

1 Culture refers to a body of information and behavioural traits that are transmitted within and between generations by social learning

vii 195 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Recovery feasibility summary

Resident Killer Whale populations are not expected to achieve high abundances that might result in a de-listing due to their ecological position as upper trophic-level predators coupled with their apparent propensity to live in relatively small populations. Despite this, and despite gaps in our knowledge, the Recovery Team views the recovery of both populations to a more robust and sustainable status as technically and biologically feasible. Both populations have males, reproductive and pre-reproductive females, and the capacity to grow. During past periods of population growth, annual increases of approximately 3% have been recorded (see 3.3.2 in Population size and trends). Growth is unlikely to exceed these levels due to the low reproductive rate of the species, and therefore the recovery of Northern and Southern Resident Killer Whales can be expected to take more than one generation. Due to its small size, the Southern Resident Killer Whale population will be particularly vulnerable to catastrophic events and continues to have a high risk of extinction during this period.

Technologies and methodologies currently exist to reduce many of the threats facing Killer Whales, their prey and their habitat. As well, identification of critical habitat and the protection of all critical habitat areas from further degradation will ensure that Resident Killer Whales have sufficient habitat for recovery. The action plan for Northern and Southern Resident Killer Whales describes 98 recovery measures to address threats to the species and monitor its recovery, many of which are underway. As Killer Whales travel regularly across international borders, it is timely that both the Washington State and the United States federal governments are also engaged in conservation actions to promote the recovery of both populations.

viii 196 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Table of contents

Preface ...... iii Acknowledgements ...... iii Strategic environmental assessment statement ...... v Executive summary ...... vi Recovery feasibility summary ...... viii 1. Introduction ...... 1 2. COSEWIC species assessment information ...... 1 3. Species information ...... 2 3.1 Description ...... 2 3.2 Distribution ...... 4 3.2.1. Global range ...... 4 3.2.2. Canadian Pacific range ...... 4 3.3 Population size and trends ...... 6 3.3.1 Global ...... 6 3.3.2 British Columbia ...... 7 3.4 Natural factors affecting population viability and recovery ...... 10 3.4.1. Biological limiting factors ...... 10 3.4.2. Other natural limiting factors ...... 15 4. Threats ...... 16 4.1. Historic threats...... 16 4.2. Current threats...... 17 4.2.1. Environmental contaminants 18 4.2.2. Reduced prey availability 23 4.2.3. Disturbance 25 4.2.4. Oil spills 35 4.2.5. Incidental mortality in fisheries 35 5. Knowledge gaps ...... 35 6. Recovery ...... 36 6.1 Recovery goal ...... 36 6.1.1 Measures of recovery success 37 6.1.2 Monitoring and research strategies 37 6.2 Recovery objectives and strategies to achieve recovery ...... 37 6.2.1 Objective 1 38 6.2.2 Objective 2 38 6.2.3 Objective 3 39 6.2.4 Objective 4 40 6.3 Effects on non-target species ...... 41 6.4 Evaluation and the status of strategies for recovery ...... 41 7. Critical habitat ...... 44 7.1 Identification of the species’ critical habitat ...... 44 7.1.1 General description of the species’ critical habitat 44 7.1.2 Information and methods used to identify critical habitat 45 7.1.3 Identification of critical habitat 47 7.2 Schedule of studies to identify critical habitat ...... 52 7.3 Activities likely to result in the destruction of critical habitat ...... 53 8. Statement on action plans ...... 58 9. References ...... 59 Appendix A. Record of cooperation and consultation ...... 77

ix 197 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Appendix B. Recovery team members ...... 79 Appendix C. Contaminant acronyms ...... 81 Appendix D. Description of critical habitat ...... 82

x 198 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

1. Introduction

Two distinct populations of Resident Killer Whales (Orcinus orca), known as the Northern and Southern Residents, occupy Canadian Pacific waters. Northern Resident Killer Whales are listed as Threatened and Southern Resident Killer Whales are listed as Endangered under the Species at Risk Act (SARA).

The Recovery Strategy for the Northern and Southern Resident Killer Whales (Orcinus orca) in Canada was finalized and published on the Species at Risk Public Registry in 2008. Minor amendments to the recovery strategy were made in 2011 to provide additional clarification regarding critical habitat for Northern and Southern Resident Killer Whales. In 2018, the recovery strategy was amended again to include identification of additional critical habitat for these populations (section 7) and to provide minor updates to background and species information. Further, minor changes were made to the formatting of this recovery strategy to adhere to current national templates where possible. This recovery strategy is considered one in a series of documents for this species that are linked and should be taken into consideration together; including the Committee on the Status of Endangered Wildlife in Canada (COSEWIC) status report (COSEWIC 2008) and an action plan (DFO 2017a).

2. COSEWIC species assessment information

Date of Assessment: November 2008

Common name: Killer Whale - Northern Resident population

Scientific name: Orcinus orca

Status: Threatened

Reason for designation: The population is small, and is limited by the availability of its principal prey, Chinook Salmon. It is also at risk from physical and acoustical disturbance, oil spills and contaminants. However, this population has been increasing slowly since monitoring began in 1975.

Occurrence: Pacific Ocean

Status history: The “North Pacific Resident populations” were given a single designation of Threatened in April 1999. Split into three populations in November 2001. The Northern Resident population was designated Threatened in November 2001. Status re-examined and confirmed in November 2008. Last assessment based on an update status report.

199 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Date of Assessment: November 2008

Common name: Killer Whale - Southern Resident population

Scientific name: Orcinus orca

Status: Endangered

Reason for designation: The population is small and declining, and the decline is expected to continue. Southern Residents are limited by the availability of their principal prey, Chinook Salmon. There are forecasts of continued low abundance of Chinook Salmon. Southern Residents are also threatened by increasing physical and acoustical disturbance, oil spills and contaminants.

Occurrence: Pacific Ocean

Status history: The “North Pacific Resident populations” were given a single designation of Threatened in April 1999. Split into three populations in November 2001. The Southern Resident population was designated Endangered in November 2001. Status re-examined and confirmed in November 2008. Last assessment based on an update status report.

3. Species information

3.1 Description

The Killer Whale is the largest member of the dolphin family, Delphinidae. Its size, striking black and white colouring and tall dorsal fin are the main identifying characteristics. Killer Whales are mainly black above and white below, with a white oval eye patch, and a grey saddle patch below the dorsal fin. Each Killer Whale has a uniquely shaped dorsal fin and saddle patch, and most animals have naturally acquired nicks and scars. Individual Killer Whales are identified using photographs of the dorsal fin, saddle patch, and sometimes eye patches (Ford et al. 2000). They are sexually dimorphic. Maximum recorded lengths and weights for male Killer Whales are 9.0 m and 5,568 kg respectively, whereas females are smaller at 7.7 m and 4,000 kg (Dahlheim and Heyning 1999). The tall triangular dorsal fin of adult males is often as high as 1.8 m, while in juveniles and adult females it reaches 0.9 m or less. In adult males, the paddle- shaped pectoral fins and tail flukes are longer and broader and the fluke tips curl downward (Bigg et al. 1987).

Currently, most authorities consider Killer Whales to be one species, Orcinus orca, having regional variations in diet, size, colouration, and vocal patterns (Heyning and Dahlheim 1988; Ford et al. 2000; Barrett-Lennard and Ellis 2001). Two and possibly three distinct species have recently been proposed for Antarctic populations (Mikhalev et al. 1981; Berzin and Vladimorov 1983; Pitman and Ensor 2003), but they are not currently widely accepted (Reeves et al. 2004). In addition, recent genetic studies report little global variation in mitochondrial DNA suggesting that the population segregation indicated by the morphological differences described above is relatively recent (Barrett-Lennard 2000; Hoelzel et al. 2002).

Three distinct forms, or ecotypes, of Killer Whale inhabit Canadian Pacific waters: Transient, Offshore, and Resident. These forms are sympatric but socially isolated and differ in their

200 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 dietary preferences, genetics, morphology and behaviour (Ford et al. 1998; 2000; Barrett- Lennard and Ellis 2001). Transient Killer Whales feed on marine mammals; particularly Harbour Seals (Phoca vitulina), porpoises, and sea lions (Ford et al. 1998). They travel in small, acoustically quiet groups that rely on stealth to find their prey (Ford and Ellis 1999). To the experienced eye, the dorsal fins of Transient Killer Whales tend to be pointed and their saddle patches are large and uniformly grey (Ford et al. 2000). Offshore Killer Whales are not as well understood as Residents and Transients. They feed primarily on elasmobranchs, but have also been documented to prey on teleost fishes, including Chinook Salmon (Heise et al. 2003; Ford et al. 2014). They often travel in large acoustically active groups of 30 or more whales, using frequent echolocation and social calls (Ford et al. 2000). The dorsal fins of Offshore Killer Whales are more rounded than those of Transients, and their saddle patches may either be uniformly grey or may contain a black region.

Resident Killer Whales are the best understood of the three ecotypes. They feed exclusively on fish and cephalopods and usually travel in acoustically active groups of 10 to 25 or more whales (Ford et al. 2000). The tips of their dorsal fins tend to be rounded at the leading edge and have a fairly abrupt angle at the trailing edge. Their saddle patches may be uniformly grey or contain a black region. The social organization of Resident Killer Whales is highly structured. Their fundamental unit is the matriline, comprising all surviving members of a female lineage. A typical matriline comprises an adult female, her offspring, and the offspring of her daughters. Both sexes remain within their natal matriline for life (Bigg et al. 1990). Social systems in which both sexes remain with their mother for life have only been described in one other mammalian species, the Long-Finned Pilot Whale (Globicephala melas) (Amos et al. 1993). Bigg et al. (1990) defined pods as groups of closely related matrilines that travel, forage, socialize, and rest with each other at least 50% of the time, and predicted that pods, like matrilines, would be stable over many generations. However, Ford and Ellis (2002) showed that inter-matriline association patterns in the Northern Residents have evolved over the past decade such that some of the pods identified by Bigg et al. (1990) now fail to meet the 50% criterion. Their analysis suggests that pods are best defined as transitional groupings that reflect the relatedness of recently diverged matrilines. Each Resident pod has a unique dialect made up of approximately a dozen discrete calls (Ford 1989; 1991). These dialects can be distinguished, providing each pod with a unique acoustic signature. Dialects are probably learned from mothers and other associated kin and are highly stable over time (Ford et al. 2000). Their function is not entirely understood, although it appears that they play an important role in mate selection (Barrett-Lennard 2000, discussed below in section 3.4.1. Culture). Despite having distinct dialects, some pods share certain calls and call variants. Pods that share one or more calls belong to a common clan.

Resident Killer Whales that share a common range and that associate at least occasionally are considered to be members of the same community or population. There are two communities of Resident Killer Whales in British Columbia, the Northern Residents and the Southern Residents. Despite having overlapping ranges, these two communities are acoustically, genetically, and culturally distinct. The Northern Resident community consists of three clans, and the Southern Resident community consists of one.

The existence of two distinct populations of Resident Killer Whales using the waters of Washington and British Columbia has been recognized by both the Canadian and U.S. governments. In 2001 COSEWIC assessed Northern Residents as Threatened and Southern Residents as Endangered. In the United States, marine mammals are afforded federal protection under both the Marine Mammal Protection Act (MMPA) and, when listed, under the

201 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Endangered Species Act (ESA). The Southern Residents were listed as ‘depleted’ under the MMPA in 2003. In February 2006, Southern Resident Killer Whales were listed as Endangered under the ESA. In June 2004, the Washington State Department of Fish and Wildlife added Southern Resident Killer Whales to their Endangered Species List.

3.2 Distribution

3.2.1. Global range

Killer Whales are found in all oceans, and are most common in areas associated with high ocean productivity in mid to high latitudes (Forney and Wade 2006). They are able to tolerate temperatures ranging from those found in polar waters to the tropics, and have been recorded in water ranging from shallow (several metres) to open ocean depths (Baird 2001).

3.2.2. Canadian Pacific range

Killer Whales are found in all three of Canada's oceans, as well as occasionally in Hudson Bay and in the Gulf of St. Lawrence. They are rarely documented in the northwestern Atlantic, but their occurrence in the eastern Arctic has been increasing in recent years (COSEWIC 2008; Ferguson et al. 2010). In British Columbia (BC), Killer Whales have been recorded throughout almost all salt-water areas, including many long inlets, narrow channels, and deep embayments (Baird 2001). The three ecotypes of BC Killer Whales (Offshore, Transient, and Resident) do not appear to interact socially despite their overlapping ranges (Ford et al. 2000). Offshore Killer Whales are most often sighted on the continental shelf off the outer coast, but they are occasionally found in protected inside waters (Ford et al. 2000). Transient Killer Whales range throughout the area, as do Resident Killer Whales (Ford and Ellis 1999; Ford et al. 2000). Residents and Transients have occasionally been seen in close proximity to each other, but rarely interact (Ford and Ellis 1999). Figure 1 shows many place names mentioned in the text, as well as the general ranges of Northern and Southern Residents.

202 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Figure 1. The coast of British Columbia and northwest Washington State showing the general ranges of Northern and Southern Resident Killer Whales.

Southern Residents

The community of Southern Residents comprises a single acoustic clan, J clan, which is composed of three pods (referred to as J, K, and L) containing a total of 20 matrilines (Ford et al. 2000). The known range of this community is from southeastern Alaska to central California (Ford et al. 2017). During summer, its members are usually found in waters off southern Vancouver Island and Northern Washington State, where they congregate to intercept migratory salmon. The main area of concentration for Southern Residents is Haro Strait and vicinity off southeastern Vancouver Island (Figure 1), but they are commonly seen in Juan de Fuca Strait, and the southern Strait of Georgia (Ford et al. 2000). Of the three Southern Resident pods, J pod is most commonly seen in inside waters throughout the year, and appears to seldom leave the Strait of Georgia-Puget Sound-Juan de Fuca Strait region in most years (Ford et al. 2000). K and L pods are more often found in western Juan de Fuca Strait and off the outer coasts of

203 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Washington State and Vancouver Island. Unlike J pod, K and L pods typically leave inshore waters in winter and return in May or June. Their range during this period is poorly known, but they have been sighted as far south as Monterey Bay, California and as far north as Chatham Strait, southeastern Alaska (Ford et al. 2017).

Northern Residents

The Northern Resident Killer Whale community comprises three acoustic clans (A, G, and R) containing 34 matrilines, which range from Glacier Bay, Alaska to Grays Harbor, Washington (Ford et al. 2000; Ford et al. 2017). From June to October, some Northern Resident Killer Whales are frequently documented in Johnstone Strait and Queen Charlotte Strait (Figure 1), off northeastern Vancouver Island (Ford et al. 2000). Their range at other times of the year is not as well understood. Small groups of Northern Residents are sometimes seen in Johnstone Strait and other inshore waters along the BC coast in winter (Ford et al. 2000) but such sightings are rare even when seasonal changes in observer effort are taken into account.

There is no evidence that clans are restricted to specific regions within the range of their community, but some show an apparent preference for particular areas (Ford et al. 2000). For example, the most commonly sighted whales off northeastern Vancouver Island belong to A clan, whereas most of the whales sighted off the west coast of Vancouver Island belong to G clan, and R clan seems to prefer the northern part of the community’s range. The range of Northern Residents overlaps with Southern Residents and with a community referred to as the southern Alaskan Residents. Northern Residents have never been seen associating with members of the Southern Resident community, and while they were observed travelling in proximity to a southern Alaskan Resident pod on one occasion (Dahlheim et al. 1997), it is not clear that social mixing took place. Genetic studies have not ruled out the possibility of occasional breeding between the Northern Resident and southern Alaskan Resident communities (Barrett-Lennard and Ellis 2001).

3.3 Population size and trends

3.3.1 Global

Little is known of the historic abundance of Killer Whales, except that they were “not numerous” (Scammon 1874). Since the early 1970s, photo-identification studies have provided reasonable population estimates for Killer Whales in the near-shore waters of the northeastern Pacific (Washington, BC, Alaska, and California), and similar work is now underway in several other coastal regions (e.g. the Gulf of California, the Russian Far East, New Zealand, Patagonia, Iceland, and Norway). In other areas line transect surveys have been used to provide population estimates. These include the Antarctic (25,000 whales, Branch and Butterworth 2001) and the Eastern Tropical Pacific (8,500 whales, Wade and Gerrodette 1993). As such, the worldwide abundance of Killer Whales is probably between 40,000 and 60,000 whales (Forney and Wade 2006). Trend data for Killer Whales are generally not available, with the exception of Killer Whales in BC (discussed below) and southern Alaskan Residents (population thought to be generally increasing, Matkin et al. 2008) and for a small population of Transients in Prince William Sound (AT1s, currently in decline, not likely to recover, Saulitis et al. 2002).

204 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

3.3.2 British Columbia

There are no population estimates for Killer Whales in BC prior to 1960. Population censuses for Killer Whales are now conducted annually using photo-identification of individuals. Population trends vary by community and clan. For the purposes of the recovery strategy, data held by the Center for Whale Research (CWR), Friday Harbor, Washington, were used to describe the population status and trends of Southern Resident Killer Whales. Data held by the Fisheries and Oceans Canada Cetacean Research Program (DFO-CRP) were used to describe the Northern Resident Killer Whale population. Whales are censused slightly differently by each research group.2

The Southern Resident count includes all whales that are seen during a calendar year, and mortalities are included in the count depending on when they take place. For example, a whale that is not seen from March onwards is assumed to be dead. There is less certainty that a whale that is not seen in November or December is dead, and it may be included in the count. In recent years, observer effort has been high and most members of the Southern Resident community are photographed annually, so the count is reasonably precise.

The Northern Resident count also includes all whales that are documented during a calendar year. However, not all members of the Northern Resident community are seen each year, so the count data are generally less precise than for the Southern Residents.

In 2017, there were a total of approximately 385 Northern and Southern Resident Killer Whales (CWR unpublished data; DFO-CRP unpublished data). By comparison there were approximately 521 Transient (Ford et al. 2013) and 300 Offshore Killer Whales (Ford et al. 2014), although these numbers are less precise than the Resident counts, because not all individuals are encountered each year (Ford et al. 2000).

Southern Residents

The size of the Southern Resident community has been known since the first complete photo- identification census in 1976 (CWR unpublished data). Figure 2 shows the size of each pod as well as the fluctuation in the total population of the Southern Resident community from 1976- 2017.

2 Note that there are small discrepancies in the Southern Resident counts in the literature due to different methods of recording when whales are considered to enter or leave the population.

205 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Figure 2. Population size and trends for Southern Resident Killer Whales from 1976-2017. Data source: Center for Whale Research (unpublished).

Although the Southern Resident community was likely increasing in size in the early 1960s, the number of whales in the community dropped dramatically in the late 1960s and early 1970s due to live capture for aquariums (Bigg and Wolman 1975). A total of 47 individuals that are known or likely to have been Southern Residents were captured and removed from the population (Bigg et al. 1990), indicating that the population was likely larger prior to these captures than in subsequent decades. The population increased 19% (3.1% per year) from a low of 70 individuals after the live-captures ended in 1973, to 83 whales in 1980, although the growth rate varied by pod (Figure 2). From 1981-1984 the population declined 11% (-2.7% per year) to 74 whales as a result of lower birth rates, higher mortality for adult females and juveniles (Taylor and Plater 2001), and lower numbers of mature animals, especially males, which was caused by selective cropping in previous years (Olesiuk et al. 1990). From 1985 to 1995, the number of Southern Residents increased by 34% (2.9% per year) to 99 animals. A surge in the number of mature individuals, an increase in births, and a decrease in deaths contributed to the population growth. Another decline began in 1996, with an extended period of poor survival (Taylor and Plater 2001; Krahn et al. 2002) and low fecundity (Krahn et al. 2004) resulting in a decline of 17% (-2.9% per year) to 81 whales in 2001. Since 2001, the population has fluctuated between 74 and 89 individuals. The population has not shown signs of recovery, and consisted of 74

206 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 members in 2018 (CWR unpublished data; Southern Resident Killer Whale Imminent Threat Assessment3).

Population viability analyses (PVA) have been used to estimate the extinction risk of Southern Resident Killer Whales (Taylor and Plater 2001; Krahn et al. 2002; 2004). As would be expected, extinction risk increases when the frequency and magnitude of catastrophes such as oil spills and disease epidemics are elevated. These models predict that if the mortality and reproductive rates of the 1990s persist, there is a 6-100% probability that the population will be extinct within 100 years, and a 68-100% risk that the population will be extinct within 300 years. Extinction of the Southern Resident population can be regarded as inevitable in these scenarios under the assumptions of the analyses, and catastrophic events simply hasten its demise. When the mortality and reproductive rates of the entire 1974-2000 period are used, the risk of the population going extinct declines to 0-55% over 100 years and 2-100% over 300 years. A more recent PVA predicted survival and recovery rates of Southern Resident Killer Whales based on sex-structured models and high-quality demographic data that encompassed one Killer Whale generation (25 years; 1987-2011). These models predicted an annual decline of 0.91% for this population, with an extinction risk of 49% over a 100-year period (Velez-Espino et al. 2014). A PVA that explored the relative importance of the primary anthropogenic threats (Chinook prey availability, noise and disturbance, and contaminants) to Southern Resident Killer Whale population trajectories was constructed by Lacy et al. (2017). These models predicted that prey limitation had the greatest potential to impact population growth, but that either higher levels of noise and disturbance or higher levels of PCB contamination would also be sufficient to shift population trajectories from slow positive growth into decline.

Northern Residents

The Northern Resident community was likely increasing in size during the early 1960s, but was cropped by the live capture fishery of 1964-1973, during which at least 14 individuals were removed. Twelve of those are known to have been from one pod (A5, Bigg et al. 1990). When first censused in 1974, the Northern Resident community was estimated to contain approximately 120 whales. Although abundance estimates for Northern Residents are less precise than those for Southern Residents because not all matrilines are seen each year, it appears that the Northern Resident community grew steadily during the period 1974 to 1991 (approximately 3.4% per year, Figure 3). The population increased to 220 animals in 1997 (growth of 3.0% per year, Towers et al. 2015). Several reasons have been postulated for the Northern Residents’ success relative to Southern Residents during this period: the population’s larger size may have buffered changes in birth and death rates, fewer animals were captured during the live-capture fishery (Olesiuk et al. 1990), and in general they are exposed to less disturbance and environmental contamination. Between 1997 and 2003, the Northern Resident community declined 7% to 205 whales in 2003 (Towers et al. 2015, Figure 3). The increased mortality rates that drove the declines during this period for both Northern and Southern Residents coincided with a period of reduced range-wide Chinook Salmon abundance (Ford et al. 2010). The Northern Resident Killer Whale population has been increasing at an annual average rate of 2.9% since that time, reaching approximately 309 individuals in 2017 (Towers et al. 2015; DFO-CRP unpublished data). Population viability analyses based on sex-structured models and high-quality demographic data predicted a 1.58% annual increase and an extinction

3 Available at: https://www.canada.ca/en/environment-climate-change/services/species-risk-public- registry/related-information/southern-resident-killer-whale-imminent-threat-assessment.html

207 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 risk of 0% for Northern Resident Killer Whales over a 100-year period (Velez-Espino et al. 2014).

Figure 3. Population size and trends for Northern Resident Killer Whales from 1974 to 2016. In years with uncertainty, the minimum and maximum population sizes are represented with grey shading. Data sources: Towers et al. (2015); DFO-CRP (unpublished).

3.4 Natural factors affecting population viability and recovery

It is important to note that Northern and Southern Resident Killer Whales have been studied primarily in protected waters during the months of May to October (Ford et al. 1998; 2000). Their behaviour and ecology in other areas and seasons is less well-known.

3.4.1. Biological limiting factors

The following description of the biology of Killer Whales is based on data from both the Northern and Southern Resident populations. Essentially, Resident Killer Whales feed on fish and do not

208 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 switch to marine mammals when their principal prey species are not abundant. They are long- lived animals with no natural predators. On average, females produce a single calf every five to six years during a 25-year reproductive period, and as a result the population has an inherently slow rate of growth. Resident Killer Whales have strong cultural traditions that influence their association and mating behaviours, which also limits the capacity for the population to grow. More detailed information on the factors that may limit the ability of Resident Killer Whale populations to grow is provided below.

Diet

Although Killer Whales feed on a wide range of prey species globally, Northern and Southern Resident Killer Whales are dietary specialists, feeding primarily on fish (Ford et al. 1998). Unlike Transient Killer Whales, Resident Killer Whales do not feed on marine mammals and the breadth of their diet appears to be quite limited. Extensive surface observations and collection of prey fragments from sites of kills by Resident Killer Whales have shown that these whales forage selectively for certain salmonids regardless of their abundance (Ford and Ellis 2005; Ford et al. 2010; Hanson et al. 2010). Chinook Salmon is the predominant prey species taken by both Northern and Southern Resident communities during May-August, but Chum Salmon is more prevalent in September-October. Coho Salmon (Oncorhynchus kisutch) are taken in low numbers in June-October, but Sockeye (O. nerka) and Pink (O. gorbuscha) Salmon are not significant prey species despite their high seasonal abundance. Non-salmonid fishes do not appear to represent an important component of Resident Killer Whale diet during May-October. Stomach content analysis from stranded Killer Whales and fecal sampling from live whales also support the importance of Chinook and Chum Salmon in Resident Killer Whale diet (Ford et al. 1998; Hanson et al. 2010).

Resident Killer Whales likely forage selectively for Chinook Salmon over other available salmonids because of the large size, high fat content, and year-round availability of this species in coastal waters (Ford et al. 1998; Ford and Ellis 2005). Killer Whales feeding at Langara Island in Haida Gwaii (Queen Charlotte Islands) are known to feed on Chinook from stocks returning to rivers as far north as northern BC and as far south as California (Ford et al. 2017). The movement patterns of Resident Killer Whales are influenced by the availability of their preferred prey. During the summer months, Resident Killer Whale distribution is associated spatially and temporally with the migratory routes of Chinook Salmon as this important prey species returns to natal streams to spawn (Ford and Ellis 2005). In fall, the presence of Chum Salmon appears to influence the movements of Resident Killer Whales. In Johnstone Strait, Chum Salmon is the primary prey species taken by Northern Residents from late September through October (Ford and Ellis 2005). Fall movements of Southern Resident pods into Puget Sound were roughly correlated with runs of Chum Salmon, as well as Chinook (Osborne 1999). Recent winter sightings of Southern Resident Killer Whales in central California were coincident with high local densities of Chinook Salmon (N. Black, Monterey Bay Whale Watch, unpublished data).

Social organization

The social structure of Killer Whales in BC appears to be complex and differs among the three ecotypes (Ford and Ellis 1999; Ford et al. 2000). The social structure of Resident Killer Whales is the best understood, and one of its unique features is that there is no permanent dispersal of either sex from the natal group. The basic social unit of Resident Killer Whales is the matriline, composed of an older female (or matriarch), her male and female offspring, and the offspring of her daughters (Ford et al. 2000). Because matriarchs have long life spans, some matrilines may

209 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 contain four or more generations. In over three decades of study, immigration and emigration have rarely been observed (Bigg et al. 1990; Ford et al. 2000). Two recent cases of juvenile Resident Killer Whales leaving their matrilines and traveling alone are considered to be exceptional, isolated incidents. One, a female calf referred to as A73, or Springer, was separated from her pod shortly after her mother died and was observed alone after a brief period of association with a pod from another clan. She was subsequently reunited with her pod and joined another matriline. The second incident involved a male calf L98, or Luna, who became isolated from his pod and all other Killer Whales for unknown reasons in 2001.

Although individuals do not disperse from their natal group, matriline splitting does occasionally occur. For example, sisters often begin to spend more and more time apart after their mother dies, and their own matrilines may eventually become socially independent (Bigg et al. 1990; Ford et al. 2000; Ford and Ellis 2002). Stredulinsky (2016) conducted an analysis of matriline fission and found that group splitting is driven primarily by population growth and by the demographic composition of groups.

Reproductive parameters

Females reach sexual maturity, defined as the age of first successful pregnancy, at 15 years on average (range 12-18 years) (Olesiuk et al. 1990). Males reach sexual maturity, defined as when the dorsal fin shape changes sufficiently to distinguish males from females, at 15 years on average (range 10-17 years). Males reach physical maturity (when the dorsal fin reaches its full height) at about 20 years. Genetic paternity testing indicates that males rarely reproduce before 25 years of age (Barrett-Lennard 2000). The gestation period of Killer Whales is typically 16 to 17 months, one of the longest of all whales (Walker et al. 1988; Duffield et al. 1995). Only single calves are normally born. Only one possible case of twins has been reported (Olesiuk et al. 1990).

Approximately equal numbers of males and females are born (Dahlheim and Heyning 1999) and newborn calves are between 218 and 257 cm long (Olesiuk et al. 1990). Haenel (1986) estimated that calves are weaned at one to two years of age. The interval between calving is usually about five years for Northern Residents and six years for Southern Residents (DFO- CRP unpublished data). However the interval is highly variable, ranges from two to 12 years, and increases with age until menopause (Olesiuk et al. 1990). Overall, females have an average of five viable calves in a 25 year reproductive lifespan (Olesiuk et al. 1990). Calving occurs year-round, but appears to peak between fall and spring.

Mating behaviour

Mating behaviour between male and female Killer Whales has rarely been observed in the wild. However, genetic evidence has revealed that Resident Killer Whales have a propensity to mate outside their matriline (and clan, in the case of Northern Residents) but inside their community (Barrett-Lennard 2000; Barrett-Lennard and Ellis 2001). This minimizes the possibility of inbreeding very effectively, but restricts the options for mating if the population becomes very small. For example, in the Southern Resident community there may be an extreme shortage of sexually mature males, particularly for L pod females, assuming females select mates outside their pod.

210 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Survival and longevity

Survival of Resident Killer Whales varies with age. Neonate mortality (from birth to six months of age) is high, reported at approximately 43% for all Residents (Olesiuk et al. 1990), and 42% for Northern Residents (Bain 1990). Accordingly, average life expectancy is reported for an animal that survives the first six months, and is estimated to be 50 years for females and 29 years for males (Olesiuk et al. 1990). Maximum longevity for females is an estimated 80-90 years and for males is 50-60 years (Olesiuk et al. 1990). Although a typical trait in most mammals, the shorter lifespan of males could be related to sexual selection (Baird 2000) or to higher levels of persistent chemicals, such as PCBs (Ross et al. 2000). The bioaccumulation of toxins is discussed in greater detail in section 4.2.1. Atypical, however, is the prolonged post reproductive period of females, discussed in the following section. Recent evidence suggests that declines in both the Northern and Southern Resident populations (all age and sex classes) can be attributed to an increase in mortality rates (Ford et al. 2005) as well as a decrease in fecundity for Southern Residents (Krahn et al. 2004). The potential causes of the population declines are discussed in section 4.

Reproductive senescence

The average life span of female Resident Killer Whales is approximately 50 years, but on average they produce their last calf at 39, and a significant number live to 70 years or more (Olesiuk et al. 1990). The ‘grandmother hypothesis’ suggests that the presence of older females in a group can increase the survival of offspring, and this may indeed be true for Killer Whales (see discussion under Culture below). In any case, when evaluating the status of Killer Whale populations, it is important to consider the age structure of the population and to note that post- reproductive adult females are no longer able to contribute directly to population growth. In an endangered population of Transient Killer Whales in southern Alaska (AT1s), no calves have been born since 1984. Since the remaining females are near or beyond their reproductive years, the population is on the verge of extinction (Saulitis et al. 2002), with virtually no prospect for recovery, even though it may persist for many more years.

Culture

Culture refers to a body of information and behavioural traits that are transmitted within and between generations by social learning. Until recently, culture was generally considered a distinguishing feature of human societies. Of late, the concept of culture has been broadened to include non-human mammals and birds (reviewed in Rendell and Whitehead 2001) and there is strong evidence for it in both Northern and Southern Resident Killer Whales, and southern Alaskan Resident Killer Whales (Ford 1991; Ford et al. 1998; Barrett-Lennard et al. 2001; Yurk et al. 2002). There is also evidence for culture in other cetaceans, such as Sperm Whales (Whitehead and Rendell 2004), although not to the same extent as for Resident Killer Whales (Rendell and Whitehead 2001).

Dialects are the best studied form of culture in Killer Whales. A calf learns its dialect from its mother and other closely related adults, retains it for life, and passes it on to the next generation with few modifications (Ford 1991; Deecke et al. 2000; Miller and Bain 2000). These culturally- transmitted dialects may play an important role in inbreeding avoidance, since females apparently prefer males from dialect groups other than their own (Barrett-Lennard 2000; Yurk et al. 2002).

211 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Culture also appears to play an important role in feeding, with dietary preferences and probably foraging techniques and areas passed on culturally (Ford et al. 1998). Culture may also select for longevity in Killer Whales, as it provides a mechanism for older individuals to increase the fitness of their offspring and relatives by transferring knowledge to them (Barrett-Lennard et al. 2001). Foster et al. (2012) found that both reproductive and post-reproductive female Resident Killer Whales increase their own offspring’s survival. This is particularly evident for older male offspring: the death of post-reproductive female Resident Killer Whales increases mortality risk by almost 14-fold in their >30 year old sons. Prey sharing among Resident Killer Whales is likely one way that older individuals increase the fitness of their offspring. Cooperative prey sharing has been documented by all age and sex classes of Resident Killer Whales, but adult females share most frequently (Wright et al. 2016). Adult female Northern Resident Killer Whales share over 90% of the fish that they catch, most often with their offspring (Ford and Ellis 2006; Wright et al. 2016), and thus play a significant role in provisioning these members of their matrilines.

Culture may help animals adapt to changing environments by allowing them to learn from each other in addition to learning from experience. For example, based on differences in foraging success by sympatric clans of Sperm Whales under different climatic regimes, Whitehead et al. (2004) suggest that cultural diversity may be even more significant than genetic diversity in helping Sperm Whales to deal with a changing ocean climate. While we do not know if this is true for Resident Killer Whales, we do know that they respond culturally to anthropogenic changes in their environment. In Alaska, Resident Killer Whales responded to longline fishing in areas of Alaska by learning to raid the gear and take fish, and this behaviour spread rapidly throughout the population (Matkin and Saulitis 1994). Depredation of Pacific Halibut longline fisheries and salmon troll fisheries is also known to occur in BC waters (Ford 2014).

Depensation

Resident Killer Whale populations are at risk simply by virtue of their small population size. In general, small populations have an increased likelihood of inbreeding and lower reproductive rates, which can lead to low genetic variability, reduced resilience against disease and pollution, reduced population fitness, and elevated extinction risks due to catastrophic events. Pacific Resident Killer Whale populations are considered small, at 74 Southern Residents in 2018 and approximately 309 Northern Residents in 2017 (CWR unpublished data; DFO-CRP unpublished data). If either Resident population is reduced further, they may be faced with a shortage of suitable mates. Among the Southern Residents, L pod females may be particularly vulnerable to this scenario because of the small number of reproductive males in J and K pod. Even under ideal conditions, the population will recover slowly because Killer Whales calve relatively infrequently.

Northern Resident Killer Whales have been shown to minimize inbreeding and its inherent risks by consistently selecting unrelated mates (Barrett-Lennard and Ellis 2001), suggesting that this population is more genetically viable than would be expected from population size alone. In contrast, recent evidence of incestuous mating among Southern Resident Killer Whales presented by Ford et al. (2018) suggests that they may be substantially more vulnerable to negative effects of inbreeding, particularly if the population remains at or below its present size for multiple generations.

Natural mortality

Killer Whales have no recorded predators, other than humans. There are several potential sources of natural mortality that may impact Killer Whales: entrapment in coastal lagoons or

212 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 constricted bays, accidental beaching, disease, parasitism, biotoxins, and starvation (Baird 2001). However, it cannot be ruled out that anthropogenic factors may make Killer Whales more vulnerable to natural sources of mortality. For example, disturbance from intense noise may cause animals to strand (Perrin and Geraci 2002). In this case, the proximate cause of death, stranding, is a natural source of mortality, but the death would be ultimately human-caused.

3.4.2. Other natural limiting factors

Entrapment and/or accidental beaching

Accidental beaching and entrapment are sometimes a source of mortality for Killer Whales. At least four mass strandings involving more than 36 individuals occurred in BC in the 1940s (Cameron 1941; Carl 1946; Pike and MacAskie 1969; Mitchell and Reeves 1988). Although the causes of mass strandings in toothed whales are uncertain, disease, parasitism, and disturbance from intense underwater noise have been suggested as possible causes (Perrin and Geraci 2002). Two possible cases of temporary entrapment have been reported for Southern Resident Killer Whales (Shore 1995; 1998). In 1991, J pod spent 11 days in Sechelt Inlet, apparently reluctant to exit through a constricted entrance with tidal rapids. In 1997, 19 Killer Whales spent 30 days in Dyes Inlet, Puget Sound, possibly because they were reluctant to pass under a noisy bridge (Shore 1998).

Disease and parasitism

Diseases in captive Killer Whales have been well studied, but little is known of diseases in wild Killer Whales (Gaydos et al. 2004). Causes of mortality for captive Killer Whales include pneumonia, systemic mycosis, other bacterial infections, and mediastinal abscesses (Greenwood and Taylor 1985). Of 16 pathogens identified in Killer Whales, three have been detected in wild individuals: marine Brucella, Edwardsiella tarda, and cetacean poxvirus (Gaydos et al. 2004). A severe infection of E. tarda resulted in the death of a Southern Resident male in 2000 (Ford et al. 2000). Marine Brucella may cause abortions and reduced fecundity in Killer Whales (Gaydos et al. 2004). Cetacean poxvirus can cause mortality in calves and causes skin lesions (Van Bressem et al. 1999). Twenty-seven additional pathogens have been identified in sympatric odontocetes that may be transmittable to Killer Whales (Gaydos et al. 2004).

External parasites of Killer Whales have been reported in Mexico (Black et al. 1997), but none have been observed on Resident Killer Whales in BC (Baird 2001). Internal parasites of Killer Whales include various trematodes, cestodes, and nematodes (Heyning and Dahlheim 1988; Raverty et al. 2014). These endoparasites are usually acquired through infected food, but the amount of infection and their contribution to Killer Whale mortality are not known at this time.

Algal blooms

Harmful algal blooms (HABs) are blooms of algae that produce biotoxins such as paralytic shellfish poison, domoic acid, saxitoxin, and brevitoxin. Such toxins can accumulate in the tissues of species that ingest them and are magnified up the food chain. Mortality of Humpback Whales (Megaptera novaeangliae) off Massachusetts in 1987 and California Sea Lions (Zalophus californianus) in California in 1998 have been linked to biotoxin exposure (Geraci et al. 1989; Scholin et al. 2000). Several species of marine mammals have been shown to have a potential susceptibility to the neurotoxic effects of biotoxins (Trainer and Baden 1999). Given the

213 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 apparent increase in HAB event frequency, and the potential for toxic effects in Killer Whales, there may be some risk to Resident Killer Whales exposed to biotoxins through HABs, although the risk is thought to be low (Krahn et al. 2002).

Regime shifts

In the North Pacific, there are widespread changes that occur in the circulation and physical properties of the ocean. These changes take place on decadal time scales and are referred to as ‘regime shifts’ (see reviews in Francis et al. 1998; Benson and Trites 2002). Such shifts may happen quite quickly, and result in dramatic changes in the distribution and/ or abundance of many species, ranging from zooplankton to fish and possibly marine mammals and seabirds. If the distribution or abundance of Resident Killer Whale prey changed significantly following a regime shift, it is possible that Killer Whales could be affected.

4. Threats 4.1. Historic threats

Pliny the Roman scholar first described a Killer Whale as an “enormous mass of flesh armed with savage teeth” during the first century AD. Since then written records have often depicted Killer Whales as savage, destructive, ferocious, and a danger to humans. However, they were rarely hunted, with the exception of Japanese, Norwegian and Russian whalers. Contemporary fishermen have viewed the Killer Whale as a competitor for their fish and a threat to their livelihood (Olesiuk et al. 1990; Ford et al. 2000). The live capture of Killer Whales for aquariums in the 1960s and early 1970s reduced local populations, some drastically.

Harvest and live captures

Killer Whales were hunted commercially, but whaling operations generally targeted other species of whales. In Canada, there are only a few harvest records of Killer Whales, most of which took place on the east coast and in the Arctic (e.g. Mitchell and Reeves 1988; Reeves and Mitchell 1988). However, large numbers of whales were taken in other areas of the world. The Japanese killed 60 Killer Whales per year between 1948 and 1957 (Nishiwaki and Handa 1958). Norwegian whalers culled 2,345 Killer Whales between 1938 and 1981 (Øien 1988). The former USSR captured approximately 25 Killer Whales per year in the Antarctic and harvested 906 whales in one season (Berzin and Vladimirov 1983). In 1982, the International Whaling Commission recommended a halt to the harvest of Killer Whales until the impact on populations was better understood. No Killer Whales have been reported taken since then, though small numbers may continue to be caught but remain unreported. For example, genetic testing has revealed the presence of Killer Whale in meat sold in Japanese and Korean markets (Baker et al. 2000).

In the late 1960s and early 1970s, Killer Whales were sought extensively for display in public aquaria. While they were captured from various areas throughout the world, the majority came from the waters of the northeastern Pacific Ocean. Between 1962 and 1974, 68 Killer Whales were taken from this area, 47 of which are known or assumed to be Southern Residents (Olesiuk et al. 1990). This cropping clearly had a major impact on the Southern Resident community, which numbered only 70 animals in 1974, and has likely affected productivity of the community for many years after the live captures ended in 1975.

214 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Intentional shootings

Historically, negative attitudes towards Killer Whales in BC led to efforts by both government and individuals to cull local populations through shooting. In 1960, the federal Fisheries Department mounted a land-based machine gun near sports fishing lodges near Campbell River to reduce the number of Killer Whales (Ford et al. 2000). Fortunately it was never fired. In the 1960s and 1970s, approximately one quarter of whales live captured for aquaria had gunshot wounds (Ford et al. 2000). Societal attitudes towards Killer Whales have changed since 1974, and fresh bullet wounds are now rarely, if ever, seen on whales in BC and Washington (Ford et al. 2000), although even occasional shootings could limit population growth.

Acoustic harassment devices

Aquaculture farms in Washington and BC have used acoustic harassment devices (AHDs) that emit loud signals underwater to reduce depredation by Harbour Seals and sea lions. Some signals may be heard from up to 50 km away (Morton and Symonds 2002). Their use at a farm near Northern Vancouver Island was associated with significant declines in the use of nearby waters by both Resident and Transient Killer Whales (Morton and Symonds 2002). Harbour Porpoise abundance was also found to drop dramatically when AHDs were in active use (Olesiuk et al. 2002). The use of AHDs is no longer permitted in BC or in Washington. They are still used at Ballard Locks in Seattle to deter sea lions, but the configuration of the canal limits the amount of noise escaping to the open ocean (Bain 1996).

4.2. Current threats

A variety of threats may directly impact Northern and Southern Resident Killer Whale populations in BC, particularly because of their small population sizes. Threats include environmental contaminants (including oil spills), reduced prey availability, disturbance and noise pollution, each of which is discussed in more detail below. An additional emerging threat, vessel strikes, was identified during a science-based review of recovery actions for Southern Resident Killer Whales (DFO 2017c). Other threats such as mortality in fishing gear have posed a threat to cetacean populations in other areas, and could potentially impact Resident Killer Whales. Climate change is affecting entire ecosystems, and it is likely that in order to survive, Killer Whales will have to adapt to the consequences of local changes in their prey base. How current threats may act synergistically to impact Killer Whales is unknown, but in other species multiple stressors have been shown to have strong negative and often lethal effects, particularly when animals carry elevated levels of environmental contaminants (Sih et al. 2004).

The extent to which Northern and Southern Resident Killer Whales are affected by anthropogenic threats varies, depending on the threat. For example, Northern Resident Killer Whales may be more vulnerable to seismic surveys on the north coast, particularly if the moratorium on oil and gas exploration is lifted, whereas Southern Residents, by virtue of the waters they spend significant time in, may be more vulnerable to environmental contaminants.

In May 2018, the Minister of Fisheries and Oceans and the Minister responsible for the Parks Canada Agency announced that they had formed the opinion, in accordance with subsection 80(2) of SARA, that Southern Resident Killer Whales face imminent threats to their survival and recovery (Southern Resident Killer Whale Imminent Threat Assessment3).

215 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

4.2.1. Environmental contaminants

There are numerous chemical and biological pollutants that may directly or indirectly impact Resident Killer Whales, ranging from persistent organic pollutants (POPs) to antibiotic resistant bacteria and exotic species. Below we describe the major types of contaminants, their sources and their potential effects on Killer Whales (where known). For a list of the acronyms mentioned below, see Appendix C. There have been only a handful of studies that have measured contaminant levels in Killer Whales, and for obvious reasons no controlled experiments have been done to assess how these contaminants may affect them directly. However, the effects of contaminants on other species such as pinnipeds are better understood, and in many cases can be generalized to Killer Whales, particularly because the physiological processes of mammals are similar across different species. This ‘weight of evidence’ approach is outlined elsewhere for marine mammals (Ross 2000).

Although it is important to assess the direct effects of contaminants, Fleeger et al. (2003) make an important case for considering their ‘indirect’ effects on community structure, as well as on individual organisms and their behaviour. In a review of 150 studies, contamination resulted in changes in species abundance and community structure. Sixty percent of the communities that were experimentally manipulated showed a reduction in upper trophic level predators, which masked, enhanced or confused the interpretation of any direct effects of contaminants on individual organisms or species.

Persistent organic pollutants (POPs)

There are likely thousands of chemicals to be found in the Killer Whales of BC, but a few key classes are of particular concern today. Studies of environmental contaminants in Resident and Transient Killer Whales in BC and Washington have revealed that they are among the most contaminated mammals in the world (Ross et al. 2000; 2002; Krahn et al. 2009). Killer Whales are vulnerable to accumulating high concentrations of POPs because they are long-lived animals that feed high in the food web (Ross et al. 2000; 2002; Rayne et al. 2004; Ross 2006). POPs are persistent, they bioaccumulate in fatty tissues, and are toxic, features that have led to increased regulatory scrutiny of these chemicals by authorities around the world. POPs include ‘legacy’ contaminants such as the polychlorinated biphenyls (PCBs) and the organochlorine pesticide DDT which are no longer widely used in industrialized countries, but remain persistent in the environment. The so-called ‘dirty dozen’ POPs are encompassed under the terms of the Stockholm Convention which aims to phase out use of chemicals of global ecotoxicological concern. They also include the polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs or furans), by-products of incomplete combustion, of pesticide manufacture, and of the (now regulated) use of elemental chlorine and pentachlorophenol (PCP) in pulp and paper bleaching and wood treatment processes, respectively. In recent years, regulations have resulted in a reduction in the release of such contaminants into the marine environment (Hagen et al. 1997).

Contaminants of ‘current concern’ in the industrial world include the new generation of polybrominated trienylethers (PBTs), flame retardants such as polybrominated diphenylethers (PBDEs), as well as currently used pesticides. Table 1 lists the POPs that are a concern for Resident Killer Whales, and the reader is referred to Grant and Ross (2002) for a more thorough synthesis of what is known about the risks that contaminants pose to Southern Resident Killer Whales.

216 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Polychlorinated biphenyls (PCBs)

Surprisingly high concentrations of PCBs are found in both Southern and Northern Resident Killer Whales relative to marine mammals from other parts of the world (Ross et al. 2000). The PCB levels found in Transients and Southern Residents exceed those found in St. Lawrence Beluga Whales (Delphinapterus leucas) by a factor of two to four times, and are considerably higher than thresholds for PCB-associated reproductive impairment, skeletal abnormalities, immunotoxicity and endocrine disruption in pinnipeds (reviewed in Ross 2000). Ross et al. (2000) found that PCB concentrations increase with age in male Killer Whales, but decline in reproductively active females. Consistent with observations in other mammals, including humans, reproductive females pass PCBs to their offspring, particularly the first born, during gestation and lactation (Tanabe and Tatsukawa 1992; Borrell et al. 1995; Ylitalo et al. 2001).

Dioxins and furans

Levels of dioxins and furans were found to be low in the blubber of Resident and Transient Killer Whale populations in BC (Ross et al. 2000). This may be partly explained by low levels of dioxins and furans in their diet, but Killer Whales may also metabolize and excrete dioxin-like compounds more effectively than PCBs (Ross 2000).

217 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Table 1. Persistent organic pollutants that may pose a risk to Resident Killer Whales.

Pollutant Use/Source Persistent Bio- Risk accumulate DDT pesticide used in some countries, banned in yes yes reproductive impairment, North America, persists in terrestrial runoff immunosuppression, (Dichlorodi- >30 years post ban, enters atmosphere from adrenal and thyroid phenyl areas where still in use effects trichloroethane PCBs electrical transformer and capacitor fluid, yes yes reproductive impairment, limited use in North America but enters skeletal abnormalities, (Polychlorinated environment from runoff, spills, and immunotoxicity, and Biphenyls) incineration endocrine disruption Dioxins and by-product of chlorine bleaching, wood yes yes thymus and liver damage, Furans product processing and incomplete birth defects, combustion. Mills less of a source now. reproductive impairment, Current sources include burning of salt-laden endocrine disruption, wood, municipal incinerators, and residential immunotoxicity, and wood and wood waste combustion; in runoff cancer from sewage sludge, wood treatment PAHs by-product of fuel combustion, aluminium yes no carcinogenic smelting, wood treatment, oil spills, (Persistent metallurgical and coking plants, pulp and Polycyclic paper mills aromatic hydrocarbons) flame retardants, flame retardants; in electrical components yes yes endocrine disruption, esp. PBBs and and backings of televisions and computers, impairs liver and thyroid PBDEs in textiles and vehicle seats, ubiquitous in environment. 2/3 product PBDEs banned in (Polybrominated Europe. Same two products withdrawn from diphenyl ethers) North American marketplace in 2005, but one (deca) product still used globally. PFOs stain, water and oil repellent (included in yes yes but in promotes tumour growth Scotchgard until recently), fire fighting foam, blood, liver, (Perfluro-octane fire retardants, insecticides and refrigerants, kidney and sulfonate) ubiquitous in environment muscle TBT, DBT antifoulant pesticide used on vessels yes yes unknown but recently associated with hearing (Tributyltin loss Dibutyltin) PCPs flame retardants, plasticizers, paints, yes yes endocrine disruption sealants and additives in lubricating oils (Polychlorinated paraffins) PCNs ship insulation, electrical wires and yes yes endocrine disruption capacitors, engine oil additive, municipal (Polychlorinated waste incineration and chlor-alkali plants, napthalenes) contaminant in PCBs APEs detergents, shampoos, paints, pesticides, moderate moderate endocrine disruption plastics, pulp and paper mills, textile industry (Alkyl-phenol found in sewage effluent and sediments ethoxylates) PCTs fire retardants, plasticizers, lubricants, inks yes yes endocrine disruption and and sealants, enters environment in runoff reproductive impairment (Polychlorinated terphenyls) References: Primarily Grant and Ross 2002, but also Lindstrom et al. 1999; Hooper and MacDonald 2000; Kannan et al. 2001; Hall et al. 2003; Van deVijver et al. 2003; Rayne et al. 2004; Song et al. 2005. 20

218 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Polybrominated diphenylethers (PBDEs)

Preliminary evidence suggests that flame retardants may be a significant and emerging concern for Resident Killer Whales (Ross 2006). Moderate levels of PBDEs were observed in 39 biopsy samples collected between 1993-1996 from Southern Resident and Transient Killer Whales, and relatively low levels were observed in Northern Residents (Rayne et al. 2004). Based on analysis of blubber samples from Harbour Seals in Puget Sound, concentrations of PBDEs doubled every 3.1 years between 1984 and 2003, but appeared to decline in 2009 (Ross et al. 2013). Regulations prohibiting the manufacture of all PBDEs in Canada came into effect in July 2009. Additionally, PBDEs have been added to the Prohibition of Toxic Substances Regulations, which prohibits all PBDEs unless present in a manufactured article. Although the toxicity of PBDEs is not well understood, they have been associated with endocrine disruption in laboratory animals (Darnerud 2003). While no conclusive link could be established as a result of the numerous other lipophilic contaminants present, PBDE concentrations were negatively associated with thyroid hormones in Grey Seals (Halichoerus grypus, Hall et al. 2003). See Ross et al. (2009) for a review of research documenting some of the sources and properties, as well as the persistence and toxicity of PBDEs.

Numerous captive and semi-field studies of pinnipeds have provided evidence that POPs affect immune function (hence, resistance to disease), hormone levels, and reproductive health (Reijnders 1986; De Swart et al. 1996; Ross 2000; Nyman et al. 2003). Using this weight of evidence as a foundation, it is not possible to ignore the substantial risks that PCBs and other POPs present to Killer Whales in the northeast Pacific. Transient Killer Whales from Prince William Sound, Alaska (AT1 population) are highly contaminated, and have had no successful reproduction since 1984, providing perhaps a population-level glimpse into the effects of high POP burdens (Ylitalo 2001). High levels of toxic chemicals may also make Killer Whales more vulnerable to disease (Ross 2002). Jepson et al. (1999) found that Harbour Porpoises that died from infectious diseases had two to three times higher concentrations of PCBs than those that died from trauma.

Biological pollutants

Biological pollution may also threaten the health of Resident Killer Whales, their habitat and their prey. These pollutants may take the form of ‘spill-over’ pathogens from human activities (e.g. pets, livestock, migrations, habitat change), virulent, antibiotic- resistant bacterial strains arising as a result of the use of antibiotics or exotic species. Emerging infectious diseases are a growing concern for marine life, as naturally occurring host-pathogen relationships are altered through human activities such as disturbance, over-fishing, habitat destruction, climate change, or pollution (Ross 2002). Killer Whales whose immune system is compromised through chemical contaminants may be increasingly vulnerable to biological pollutants. Although no disease-related mass mortalities have been observed among BC’s marine mammals, Morbillivirus has been detected in marine-dwelling River Otters (Mos et al. 2003), highlighting the potential risk of this or related pathogens to Killer Whales. In other areas, Morbillivirus outbreaks have caused mass mortalities of seals (Grachev et al. 1989; Kennedy et al. 2000) and dolphins (Aguilar and Borrell 1994). Pathogens such as Morbillivirus are capable of spreading extremely quickly (3,000 km/year), likely because in the marine environment there are few barriers to dispersal (McCallum et al. 2003).

The introduction of exotic species has changed habitats in other areas (e.g. Zebra Mussels in the Great Lakes, Eurasian Milfoil into freshwater lakes) and introduced species have the potential to impact local ecosystems here. In BC, Atlantic Salmon that have escaped from

219 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 aquaculture operations have successfully spawned in freshwater (Volpe et al. 2000). The extent to which this is occurring and how Atlantic Salmon would compete with Pacific salmon, the preferred prey of Residents (Ford et al. 1998), is not well known at this time.

Trace metals

Trace metals occur naturally in the marine environment, but elevated concentrations sufficient to be a concern to marine mammals may be found in localized areas such as urban and industrial centres (Grant and Ross 2002). Some, such as cadmium, mercury, copper and lead may have toxic effects even at relatively low concentrations, and could impact Killer Whales, although effects on their prey and/ or habitat are more likely.

Little information is available on the levels and effects of trace metals on marine mammals in the Pacific. However, in a small sample of stranded Killer Whales, Residents showed higher levels of mercury than Transients (Langelier et al. 1990). In the western Pacific, all odontocete meat sampled from Japanese markets contained amounts of mercury that exceeded the level permitted for human consumption (Endo et al. 2003). However, the historical exposure of high trophic level marine mammals to naturally elevated concentrations of mercury in prey has resulted in their evolved ability to detoxify this toxic metal through the formation of mercury- selenium crystals in the liver (Martoja and Berry 1980).

Sources of contaminants

Monitoring the sources and levels of environmental contaminants is particularly challenging given that each year, up to 1,000 new chemicals are released into the environment globally (Haggarty et al. 2003). The high contaminant levels found in Southern Residents may arise from consuming prey that are from industrialized areas near the BC-Washington border, which are more contaminated than the prey of Northern Residents (Ross et al. 2000; Cullon et al. 2009). In Japan, odontocetes that travelled in more industrialized areas carried higher contaminant loads than those found in more remote areas (Endo et al. 2003). In a study of Harbour Seals in BC and Washington, Ross et al. (2004) found that although PCB levels were a concern in all areas, seals from Puget Sound were seven times more PCB-contaminated than were seals from the Strait of Georgia. This suggests that the food web within Puget Sound has been contaminated with PCBs, such that Killer Whales consuming prey items from this region may be vulnerable to increased contaminant exposure. Chinook Salmon, one of the Resident Killer Whales’ preferred prey species (Ford et al. 1998; Ford and Ellis 2005), feed in the upper trophic levels in the food web, and those from Puget Sound are relatively contaminated with PCBs (O’Neill et al. 1998; Cullon et al. 2009). Studies suggest that most salmonids are ‘importing’ contaminants from their time at sea, reflecting global environmental contamination (O’Neill et al. 1998; Ewald et al. 1998).

Although DDT was banned in Canada in 1989 and over 40 years ago in the United States, it continues to enter the ocean from terrestrial runoff (Hartwell 2004) as well as from atmospheric transport from countries where it is still in use. Dioxins (PCDDs) and furans (PCDFs) represent highly toxic by-products of chlorine bleaching and associated wood treatment, and incomplete combustion. Source controls and regulations have greatly reduced their input in to the coastal environments of BC and Washington.

220 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Contaminants enter the marine environment from local, regional and international sources. These are discussed in detail in Haggarty et al. (2003). Local point sources of contaminants into the marine environment include:  pulp and paper mills  wood treatment facilities  municipal effluent outfalls  petrochemical facilities and  mines

Indirect sources (non-point source pollutants) include:  sewer overflows (e.g. organic wastes, household products, pharmaceuticals and personal care products)  urban runoff and storm-water drainage (e.g. pesticides, metals, hydrocarbons, herbicides and animal wastes)  agriculture (e.g. pesticides, herbicides, animal wastes and antibiotics)  forestry (e.g. pesticides, herbicides, fire-control chemicals, anti-sapstain chemicals, log booms and storage areas) and  aquaculture (e.g. organic wastes, chemical contaminants [antibiotics, feed additives, pharmaceuticals, pesticides and antifouling on nets])

Garrett and Ross (2010) describe the Canadian and U.S. federal, provincial and state agencies responsible for the monitoring, mitigation and regulation of environmental contaminants and their sources.

Shipping also represents a risk to the ecological integrity of coastal regions. Both intentional and unintentional discharge of chemicals and biological waste are added sources of pollution in all coastal areas, but particularly in high traffic zones. In addition, the introduction of exotic and invasive species carried on ship hulls and in ballast water have the potential to dramatically alter the habitats they have colonized (e.g. European Green Crabs, Zebra Mussels, the alga Caulerpa taxifola). Numerous invasive invertebrates have been found in the ballast water of ships at anchor in Vancouver Harbour (Levings et al. 2004), although the ecological significance of such introductions is unclear.

In addition, some pollutants such as PCBs, DDT and other chemicals are transported through atmospheric processes and ocean currents, and may travel to the west coast of North America from as far away as Asia in less than 5-8 days (Wilkening et al. 2000). Consequently, the northeastern Pacific may be a sink for globally produced POPs (Ross et al. 2000; 2004; 2006).

Certain ‘legacy’ POPs such as PCBs and DDT have been phased out of industrialized countries and their concentrations are slowly decreasing in the marine environment (Muir et al. 1999), although these declines have levelled off (Addison and Stobo 2001). However, levels of other ‘new’ POPs such as deca-PBDEs continue to increase globally, and represent the PCBs of the future (Ross 2006; Law et al. 2014). Unlike PCBs, which were generally used in a limited range of applications such as electrical transformers and capacitors, PBDEs have been widely used in many industrial and consumer applications and are incorporated into plastics, textiles and foam.

4.2.2. Reduced prey availability

Answering the question as to whether Killer Whales may be prey limited is complex. While the year-round diet of Resident Killer Whales is not well known, at certain times of the year salmon,

221 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 particularly Chinook and Chum, are known to be important prey (see section 3.4.1. Diet). Ford et al. (2005) found that trends in the mortality rates of Southern and Northern Resident Killer Whales were correlated with each other, and that both were strongly related to fluctuations in the abundance of Chinook Salmon, but not Chum Salmon. Birth rates were also correlated with Chinook Salmon abundance, but more weakly than mortalities.

Less is known about the prey of Resident Killer Whales and their distribution and abundance during the months of November to April. This is due to the inherent challenges of studying whales during the winter months, and because the whales move from inshore areas where they are more concentrated during summer and range widely along the coast during the winter and early spring. Thus when considering the availability of prey to Resident Killer Whales, it should be noted that we have limited knowledge of what other prey species may be important to them, and the discussion below focuses on species that are known to be important.

Changes in salmon abundance and availability

Assessing the status of salmon stocks and their availability to Resident Killer Whales is challenging to interpret and often fraught with controversy. Until the middle of the 20th century, many wild salmon stocks experienced significant declines due to overfishing, habitat degradation, restrictions in access to spawning grounds due to landslides, and changes in ocean productivity (summarized in Krahn et al. 2002 and Wiles 2004). The situation changed between 1975 and 1993, and the total abundance of North Pacific salmon doubled (Bigler et al. 1996) due to hatchery enhancement, changes in fisheries management practices and a favourable climatic regime (Bigler et al. 1996; Beamish et al. 1997). Since the early 1990s many of these stocks have declined in number and specific causes have not been identified. Some studies have questioned the role of enhancement (Beamish et al. 1997, and reviewed in Gardner et al. 2004) but other potential problems such as marine survival appear to be a factor. At present 28 of 52 different wild Pacific salmon stocks in the lower 48 states of the U.S. are considered at risk under the U.S. ESA (NOAA 2017). In southwestern BC by 1990, salmon from one-third of the spawning rivers had been lost or were seriously depleted (Riddell 1993). Recognizing that many salmon stocks are under threat, Fisheries and Oceans Canada announced a new Wild Salmon Policy (WSP) in December 2004 (DFO 2005), designed to restore and maintain healthy and diverse wild salmon populations and their habitat. Since 2005, DFO has used the WSP to guide its work toward restoring healthy salmon populations, and development of a detailed implementation plan for the WSP is underway. If these and other actions are successful, salmon may gradually become more available to Resident Killer Whales.

Resident Killer Whales tend to be found in high concentrations in specific areas during the period when salmon are returning to rivers to spawn. This likely reflects the fact that salmon are not as widely dispersed at this time as they are during the rest of their life cycle. There is a great deal of diversity in the timing of the spawning period for salmon. For example, the Upper Columbia River has a spring run and a summer/fall run of Chinook. These runs are considered distinct stocks because they do not interbreed. The spring run is Endangered under the ESA in the U.S., yet the summer/ fall run is not at risk (NOAA 2017). This illustrates the need to consider the timing of the spawning period of each salmon stock when assessing the availability of salmon for Killer Whales, in order to ensure an adequate year-round food supply. Chinook Salmon are longer lived than other salmon species and spawn at different ages (Healey 1991). It is likely that their year-round availability in nearshore waters is a key factor, along with body size and lipid content, in Chinook being the preferred salmonid prey of Resident Killer Whales (Ford and Ellis 2005).

222 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

To address the scientific uncertainty regarding the impact of sea lice on salmon, and the relationship of this to Killer Whales, DFO and others are conducting scientific research to assess and protect the health of the wild Pink and Chum Salmon resource in the Broughton Archipelago.

Depressed Chinook stocks

Chinook Salmon, the principal prey of BC’s Resident Killer Whales, is one of the least abundant species of salmon in BC (Riddell 2004). However, unlike other salmon, many populations of Chinook remain in nearshore waters during the ocean phase of their life cycle. As a result they are available on a more year-round basis to Killer Whales, but are also more vulnerable to pollution (discussed in section 4.2.1 Environmental contaminants).

Chinook abundance dropped in the 1970s and 1980s, but escapements increased until the early 1990s in some rivers, primarily due to hatchery production (Beamish et al. 1997). In Washington, hatchery fish now account for about 75% of all harvested Chinook (Mahnken et al. 1998 in Wiles 2004). In un-enhanced river systems in central and northern BC, Chinook numbers remain depressed (Riddell 2004) and nine of 17 Chinook stocks in Washington, Oregon and California are listed under the ESA (NOAA 2017). It is likely that Chinook is an important limiting factor in the population dynamics of Resident Killer Whales (Ford et al. 2005; Ward et al. 2009; Ford et al. 2010). This may explain why Southern Resident Killer Whales have appeared in places as distant as off the Columbia River and northern California to the south and off southeastern Alaska in the north (Ford et al. 2017). Their presence was associated with unusually large returns of Chinook Salmon, which they may have had to seek out because of less abundant prey within their traditional range. When prey availability is reduced, Killer Whales may be forced to spend more time and travel greater distances to forage for their food, or switch to less profitable prey, which could lead to lower reproductive rates and higher mortality rates.

In addition to reduced Chinook abundance, the quality of individual fish appears also to have declined over recent decades. Average weights of Chinook Salmon in nine populations from BC to California declined by up to 45% between 1975 and 1993 (Bigler et al. 1996). Thus, the nutritional yield of each Chinook Salmon may be significantly less today than it was in past years, which may have an impact on the overall foraging energetics of Resident Killer Whales.

4.2.3. Disturbance

All cetaceans, including Resident Killer Whales, are being subjected to increasing amounts of disturbance from vessels, aircraft, and anthropogenic noise (IWC 2004). Both private and commercial vessel traffic have increased dramatically in recent years, and Killer Whales must navigate in increasingly busy waters (Osborne 1999; Foote et al. 2004). Industrial activities such as dredging, drilling, construction, seismic testing and military sonar, and other vessel use of low and mid-frequency sonars also impact the acoustic environment (Richardson et al. 1995; NRC 2003). The means by which physical and/or acoustic disturbance can affect Resident Killer Whales at both the individual and population level are not well understood, but may depend on whether the disturbance is chronic (such as whale watching) or acute (such as seismic surveys). Other factors, including the animal’s condition, previous exposure (potentially causing sensitization or habituation), age, sex, and behavioural state also influence how disturbance affects whales (e.g. Williams et al. 2014). In addition, environmental factors, such as El Niño events that may change the availability of prey, may make animals more vulnerable to disruption than they would be otherwise. The sources of both physical and acoustic disturbance and their potential impact on Resident Killer Whales are discussed in greater detail below.

223 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

A current challenge in studying the effects of disturbance is in finding informative ways to describe and measure them, and to date the question of whether a source of disturbance is likely to result in effects at the population level can be difficult to answer. Responses to disturbance may range from slight differences in surfacing and breathing rates to active avoidance of an area. Even if the disturbance causes immediate death, carcasses are rarely recovered (regardless of the cause of death, only 6% of Killer Whale carcasses are recovered, DFO-CRP unpublished data). As well, animals may show no obvious behavioural responses to disturbance, yet still be negatively affected. For example, Todd et al. (1996) found that Humpback Whales remained in the vicinity of underwater explosions, and showed no obvious behavioural responses to them. However they experienced significantly higher entanglement rates during this time, and necropsies of two whales that drowned in nets revealed acoustic trauma (Ketten et al. 1993). Thus a lack of a measurable behavioural response to a stimulus does not necessarily imply the disturbance does not have negative consequences. A parallel may exist with humans, since people exposed to chronic noise lose their hearing more quickly than those that are not exposed to chronic noise. The consequences of hearing loss for cetaceans are likely fatal.

Measures for changes in behaviour may also not be subtle enough to detect disturbance. Whitehead (2003) re-analyzed data that were reported to indicate that Sperm Whales did not show behavioural responses to surveys using high-intensity sound. He segregated the responses according to whale density in the area and found that contrary to earlier conclusions, when whale density was low, Sperm Whales avoided seismic activity. When densities were high, whales remained in the vicinity. He suggested that whales may have been reluctant to leave a rich feeding area despite the disturbance.

Whale watching

Commercial whale watching has grown dramatically in BC, with just a few boats carrying less than 1,000 passengers per year in the late 1970s and early 1980s to 80 boats carrying half a million passengers per year in 1998 (Osborne 1991; Baird 2002; Osborne et al. 2003). In 2015, there were 93 active commercial whale watching vessels in the Salish Sea alone (Seely et al. 2017). Whale watchers tend to focus on Resident Killer Whales in their most predictable locations, Haro Strait and Johnstone Strait. Vessels in the vicinity of whales include privately owned kayaks, sailboats and powerboats as well as commercial whale watch vessels. While the benefits of public education and increased awareness that can be achieved through guided whale watching are well established, concern over the effects of whale watching on Killer Whales has grown with the industry itself. This concern has prompted the development of industry initiated viewing guidelines and has resulted in studies that have attempted to measure responses of the whales to such focused attention (e.g. Kruse 1991; Williams et al. 2002a; b; Williams et al. 2014), as well as the behaviour of boaters around whales (e.g. Jelinski et al. 2002). Whale watching activities have the potential to disturb marine mammals through both the physical presence and activity of boats, as well as the increased underwater noise levels boat engines generate.

Under the Fisheries Act in Canada and the MMPA in the U.S., disturbance (harassment) of marine mammals, including Killer Whales, is prohibited. No special provisions or exemptions to this prohibition have been made for commercial whale watch operators and the commercial fleet is subject to the same regulatory restrictions as recreational boaters. Voluntary guidelines for

224 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 viewing marine wildlife4 were developed by DFO, the National Marine Fisheries Service (NMFS), and collaborators in 2002 to protect marine mammals, including Resident Killer Whales, from disturbance. These guidelines are reviewed and revised periodically. Additionally, industry associations, including the Pacific Whale Watch Association (PWWA) and the North Island Marine Mammal Stewardship Association (NIMMSA) have developed codes of conduct for marine mammal viewing by member businesses (NIMMSA 2016; PWWA 2017).

In 2011, the U.S. National Oceanic and Atmospheric Administration (NOAA) adopted federal vessel regulations to prohibit vessels from approaching Killer Whales within 200 yards (183 metres) and from parking in the path within 400 yards (366 metres) of Killer Whales. These regulations apply to all vessels in inland waters of Washington State, with exemptions for vessels that are actively engaged in commercial or First Nations fisheries, for research vessels under permit, ships in shipping lanes and government vessels in the course of official duties. Effective July 11, 2018, amendments to the Canadian Marine Mammal Regulations under the Fisheries Act include a minimum approach distance of 100 metres for most whales, dolphins and porpoises to legally protect these animals from human disturbances as well as a new mandatory requirement for all marine vessels (including recreational boats) to stay at least 200 meters away from Killer Whales in BC and the Pacific Ocean (DFO 2018).

There are several projects that focus on educating the boating public both on and off the water about appropriate conduct in the vicinity of marine mammals. They also monitor vessel activity in the presence of whales. Current projects include the Soundwatch Boater Education Program in the San Juan Islands; Straitwatch in adjoining Canadian waters (Haro and Juan de Fuca Straits), Johnstone Strait, and occasionally off the west coast of Vancouver Island; and the Robson Bight Marine Warden Program in Johnstone Strait, while past projects include the Marine Mammal Monitoring Project in Victoria, BC. All of these programs are run by non-profit organizations that do not have guaranteed funding. Smith and Bain (2002) found that commercial operators increased their compliance with a voluntary 0.4 km ‘no boat’ zone in the San Juan Islands from less than 80% to over 90% when Soundwatch was present on the water.

Boat activity has been linked to short-term behavioural changes in Resident Killer Whales (Kruse 1991; Smith and Bain 2002; Williams et al. 2002a; 2002b). They have been known to swim faster, travel in less predictable paths, alter dive lengths, move into open water, and alter normal behaviour patterns at the surface in response to vessel presence (Kruse 1991; Williams et al. 2002a; 2002b). Foote et al. (2004) found that Southern Resident Killer Whales significantly increased the duration of their calls when boats were present, and suggested that this was an adaptation to the masking effects caused by increased noise levels. Additionally, Holt et al. (2009) found that Killer Whales increased the amplitude of their calls in response to increased vessel noise.

Although studies have shown short-term responses of Killer Whales to whale watching vessels, the long-term effects of whale watching on the health of Killer Whale populations are not known (Trites et al. 2002). Increased whale watching operations between the mid-1980s and 2001 may have resulted in a potential 20% increase in energetic expenditures of Killer Whales due to increased swimming velocity (Kriete 1995; 2002). Bain (2002) found that although the decline of Southern Residents followed the increase in commercial whale watching, the relationship was much more complex. He suggested that other variables, such as changes in the availability of

4 Be Whale Wise Guidelines for marine wildlife viewing are available at: http://www.bewhalewise.org/marine-wildlife-guidelines/

225 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 prey, were also likely significant. Whether whale watching is a significant threat to Killer Whales or not, both the Northern and Southern Resident populations continue to return to their traditional summer ranges despite increased whale watching activity. This may reflect their strong cultural behaviours or the distribution of their prey.

Underwater noise

At the time the COSEWIC status report on Killer Whales was written (Baird 2001), relatively little was known about the effects of underwater noise on marine mammals. Previous research had focused primarily on powerful noise sources with the potential to cause immediate injury or death, rather than chronic lower level noise sources (Richardson et al. 1995). Since then, there has been a rapidly growing awareness that noise is a significant threat that degrades habitat and adversely affects marine life (IUCN 2004; IWC 2004). It is estimated that ambient (background) underwater noise levels have increased an average of 15 dB in the past 50 years throughout the world’s oceans (NRC 2003).

Killer Whales have evolved in the underwater darkness using sound much the way terrestrial animals use vision: to detect prey, to communicate and to acquire information about their environment. Anthropogenic noise can interfere with all these activities in critically important ways, such as disrupting communication, reducing the distance over which social groups can detect each other, masking echolocation and hence reducing the distance over which the animals can detect their prey, potentially displacing them from preferred feeding habitats, displacing prey, impairing hearing, either temporarily or permanently and in extreme cases causing death (Bain and Dahlheim 1994; Barrett-Lennard et al. 1996; Bain 2002; Erbe 2002; NRC 2003; Au et al. 2004).

The challenges of using and interpreting behavioural responses of marine mammals to noise as a measure of disturbance are discussed above. Opportunities to measure physiological responses to anthropogenic noise are much rarer, but provide insight into the mechanisms by which noise could impact animals at the individual, and potentially population level. Physiological responses to anthropogenic noise that have been measured in marine mammals include both temporary and permanent hearing threshold shifts, the production of stress hormones and tissue damage, likely due to air bubble formation or as a result of resonance phenomena (Ketten et al. 1993; Crum and Mao 1996; Evans and England 2001; Finneran 2003; Jepson et al. 2003; Fernandez et al. 2004). Marine mammals, including Killer Whales, may be particularly vulnerable to resonance because of the air-filled cavities in their sinuses and middle ear, their lungs and small gas bubbles in their bowels. While the mechanism by which high- intensity sound can cause lethal and sub-lethal effects on cetaceans is not completely understood (Fernandez et al. 2004; Piantadosi and Thalmann 2004), loud anthropogenic sources of noise, particularly low and mid-frequency military sonars, have been implicated in mass stranding and mortality events around the world, and the subject urgently merits further study. Animals already affected by anthropogenic stressors such as environmental contaminants may be particularly vulnerable to additional stresses such as noise (Sih et al. 2004).

Sounds travel as waves much more quickly through water than air (1,530 vs. 340 m/s). The perceptual features of sound, “pitch” and “loudness,” have physical analogues. How high or low- pitched a sound is can be described in terms of its frequency, and is measured in hertz (Hz). Human hearing ranges from approximately 20 to 20,000 Hz (20 kHz), and is best between 600 and 2,000 Hz. The peak hearing sensitivity of Killer Whales is at approximately 20 kHz, although they show behavioural responses to sound from 75 Hz to over 100 kHz (Hall and Johnson 1972;

226 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Syzmanski et al. 1999). Killer Whale calls contain energy throughout this frequency range, and many echolocation clicks are centered at 20 kHz.

The ‘loudness’ of a sound is described in terms of its pressure. For the purposes of consistency, the units of measure used here are dB RMS re 1 Pa. By convention, noise sources are compared in terms of their “source levels” by estimating the level that would be measured at 1 m from the underwater sound source. In general, the further away from a sound source, the quieter the received sound level, although physical and oceanographic features of the marine environment can affect how quickly a sound attenuates (gets quieter). High frequency sounds attenuate much more rapidly than low frequency sounds under uniform conditions in the open ocean, but a number of factors influence sound propagation and high frequencies may propagate further than low frequencies in shallow water or places with complex bottom terrain. Temperature, salinity, depth, bottom topography and other physical factors must all be taken into account to accurately predict the intensity of sound reaching a whale.

The characteristics of some underwater noise sources are briefly described in Table 2. It is important to consider the length of time that animals are exposed to sounds, and sound loudness and frequency. As well, some sounds are continuous, whereas others are pulses of sound that are generated intermittently. The frequency composition also varies, ranging from broadband sounds such as seismic surveys, to narrowband sounds such as military sonar that are only broadcast across a limited range of frequencies.

Sounds at received levels of 120 dB typically disrupt the behaviour of 50% of exposed cetaceans (Richardson et al. 1995). Williams et al. (2002) found behavioural changes in Northern Residents at received levels estimated at about 105-110 dB. However, with increasing use of loud, low frequency noise in activities such as ocean acoustic tomography and low frequency active sonar, which are detectable at ranges of thousands of kilometres, there has been pressure to raise the threshold for regulatory intervention. In the United States, NMFS recently updated guidance on underwater acoustic threshold levels for avoiding permanent hearing threshold shifts (PTS) in marine mammals (NMFS 2016). This guidance expressed thresholds for hearing shifts for impulsive sounds in terms of both the cumulative sound exposure level and peak sound pressure level. Onset of PTS is considered to occur when either of the two metrics is exceeded. For cetaceans in the mid-frequency hearing group, including Killer Whales, the PTS acoustic thresholds (received levels) for impulsive sounds were identified as 230 dB for the peak sound pressure level and 185 dB for the cumulative sound exposure level. For non-impulsive sounds, the PTS onset acoustic threshold for mid-frequency cetaceans is 198 dB. It should be noted that these acoustic thresholds are just one tool for determining and mitigating the impacts of sound exposure on marine mammals. Behavioural impact thresholds and auditory masking assessments should also be considered (NMFS 2016).

227 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Table 2. Signal structure, frequency range and source levels of anthropogenic noise. Modified from Table 2-1b in NRC (2003) and Table 6.8 in Richardson et al. (1995). Source Signal Structure Frequency Range Source Level (dB re 1 Pa at 1 m) Seismic surveys impulsive broadband >240 >0 Hz to >100kHz

Military Sonar surveillance pulsed tones <1kHz >230 tactical pulsed tones >1kHz to < 10kHZ 200 to 235+ weapon/ counter pulsed tones and >10kHz to 100kHz 190 to 220 weapon wideband pulses Construction broadband and tones <10kHz to 10+kHz NA Dredging broadband and tones <10Hz to <10kHz NA Explosions impulsive broadband >240 Commercial shipping continuous 10Hz to >1kHz 160 to 200 Commercial sonars pulsed tones 28kHz to >200kHz 160 to 210

Military sonar

Military active sonar is used in military operations for target detection, localization and classification (NRC 2003). Unlike passive sonar systems, which listen for sounds, active sonar units transmit pulses of tones at frequencies from <1 to >100 kHz and source levels of 200-235 (or more) dB re 1 µPa at 1 m depending on the application (Evans and England 2001). There is now a growing weight of evidence that these sources of underwater noise may pose a significant threat to cetaceans. Active military sonar has been associated with increased strandings of beaked whales and Humpback Whales (numerous incidents summarized in IWC 2004). In October 2004, the European Parliament called on its member nations to suspend the use of all high-intensity military sonar until further research can determine what effects it may have on marine life (European Parliament Resolution P6 TA, 2004).

For security reasons, information on the specifications of military active sonar is difficult to obtain, and much of what is available is based on U.S. Navy equipment. Given that the U.S. Navy engages in joint operations with the Canadian military in both the Strait of Georgia and off the west coast of Vancouver Island, and that both Northern and Southern Resident Killer Whales travel in U.S. waters, the threat that active sonar may pose must be considered and precautionary measures should be considered by both navies. Southern Resident Killer Whales may be especially vulnerable because they spend significant time in the waters of Washington State, where a large naval exercise area runs parallel to the coast.

Military active sonars may be categorized as: surveillance (low frequency, <1 kHz), tactical (mid frequency, 1 to 10 kHz), and weapon/counter weapon (high frequency, >10 - 100 kHz, see Table 2). Tactical sonars can have detection ranges of 10s of kms, and surveillance low frequency active sonars can be detected at ranges of 100s of kms (NRC 2003). The use of SURTASS (Surveillance Towed Active Sensor System) LFA (Low Frequency Active) sonar has been controversial because of concerns about its potential effects on marine life (EIS 2007).

The Canadian Department of National Defence’s Research Agency (DRDC) conducted research to investigate low frequency active tactical sonar through the Towed Integrated Active Passive Sonar (TIAPS) off the Atlantic Coast (Bottomely and Theriault 2003). The maximum

228 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 source level of the TIAPS system was 223 dB re 1Pa @ 1m (Theriault pers. comm. 2007). Mitigation measures were applied (see Bottomely and Theriault 2003 for details) and no incidents involving marine mammals were reported. There are no plans to acquire this particular sonar for Canadian military use, and present defence policy requires that any future acquisition and testing of sonar systems will include environmental considerations (Freeman pers. comm. 2007).

Mid-frequency tactical sonar systems operating at 1-10 kHz are used to detect mines and submarines. They have been associated with mass stranding events in the Bahamas, Canary Islands, Greece and the Gulf of California (IWC 2004). Mid-frequency sonar exercises conducted by the USS Shoup on May 5, 2003 in Haro Strait were reported to correspond with changes in behaviour in members of J pod that were foraging 47 km away at the time, and resulted in behaviour more extreme than observed in response to any other disturbance. The pod was observed trying to leave the area while the ship was 22 km away and ultimately pod members separated and left the area in different directions when the USS Shoup passed by at a range of 3 km (D. Bain, personal observation and personal communication; K.C. Balcomb, in Wiles 2004). Up to 100 Dall’s Porpoises and a Minke Whale were also seen leaving the area at high speed. Extensive examination of the 11 concurrent Harbour Porpoise strandings found no definitive signs of acoustic trauma, but the cause of death could not be determined for six animals, and the possibility of acoustic trauma as a contributory factor in the deaths of the remaining five porpoises could not be ruled out (lesions consistent with both acoustic trauma and alternative explanations were observed; NMFS 2004). Further, all members of J pod were still alive more than two years after the incident.

The Canadian Navy has five principal types of military sonar emitters. The SQS 510 sonar is the primary mid-frequency sonar used for anti-submarine search and is the most powerful. It is currently fitted to six ships on the west coast. In comparison, the U.S. Navy’s SQS 53C sonar, such as that used on the USS Shoup, emits ten times more energy than the Canadian 510 sonar. The Canadian Navy also uses helicopter dipping sonars and active sonobuoys, though these emit far less energy than the 510 (D. Freeman, Department of National Defence, pers. comm. 2007).

The Canadian Navy uses active sonar during training exercises and equipment testing in designated training areas. However, sonar operations may also take place in other waters along the Pacific coast. To mitigate the potential impacts of sonar use, Department of National Defence (DND) ship personnel receive training in marine mammal identification and detection. The current Maritime Command Order 46-13 for marine mammal mitigation is to avoid transmission of sonar any time a marine mammal is observed within the defined mitigation avoidance zone specific to each type of sonar. These zones were determined using the interim NMFS thresholds for potential behavioural disturbance (160 dB) and physical injury (180 dB) (Freeman pers. comm. 2007). Concerns remain that some impacts may occur beyond the visible horizon, and these will be difficult or impossible to observe or mitigate.

Canadian test ranges are also used by other navies to test equipment and train personnel. They follow Canadian procedures for use of these ranges, which includes marine mammal impact assessment and mitigation (Freeman pers. comm. 2005). When conducting joint exercises in Canadian waters, other navies are provided direction including sonar mitigation protocols, prior to and during exercises. As little is known about the offshore distribution of Resident Killer Whales, especially during the winter months, they may be vulnerable to the use of sonar in the offshore ranges. There are no military active sonar exercise ranges within the critical habitat areas that have been identified to date.

229 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Seismic surveys

Airguns are used in geophysical surveys and to detect and monitor earthquake faults and other structures such as oil and gas deposits beneath the sea floor. The following information on the characteristics of seismic surveys comes from NRC (2003) unless mentioned otherwise. Like military sonar, seismic surveys generate high intensity sounds. Most of their energy is concentrated at frequencies between 5-300 Hz and maximum pressure levels of 260 dB re 1Pa at 1 m. However, unlike military sonars, airgun arrays used for seismic surveys generate broadband noise that extends to over 100 kHz (Calambokidis et al. 1998).

Current survey methods use one or more airguns that are towed behind a ship. Airgun arrays range in size from 2,000-8,000 cu in, depending on the application. The pulses of noise fired from these guns penetrate the seafloor surface for distances of up to 10 km deep. The arrays are towed at approximately 2.6 m/s (5 knots) and the airguns are fired every 10-12 seconds. The question of whether Killer Whales could sustain swimming the long distance necessary to avoid these sound sources needs to be addressed. Seismic surveys using powerful airgun arrays have been detected at distances of over 3,000 km from their source (Niekurk et al. 2004).

DFO receives occasional applications for permits for geophysical surveys from industry, government agencies such as Natural Resources Canada, and from universities. Currently, however, a moratorium on offshore oil and gas exploration in BC remains in place. As awareness is growing on the potential threats of high intensity sound on marine life (IUCN 2004; IWC 2004), the potential impacts of broadband high energy noise on Killer Whales must be considered. DFO has developed the Statement of Canadian Practice with respect to the Mitigation of Seismic Sound in the Marine Environment, which is reviewed annually to allow for revisions to reflect new technologies, scientific findings, and industry practices (DFO 2016a). In the Pacific Region, each proposed seismic survey is reviewed and case by case mitigation measures are developed based on the species of concern in the area of the survey.

Systematic observations of cetaceans during seismic surveys have been carried out in UK waters, and have shown that Killer Whales and other cetaceans were generally seen further away during periods when airgun arrays were firing (Stone 2003). Behavioural studies in other areas have shown mixed responses to seismic surveys. Grey and Bowhead Whales appeared to avoid seismic surveys (Malme and Miles 1987; Ljungblad et al. 1988; Myrberg 1990). Male Sperm Whales and feeding Humpback Whales did not avoid seismic surveys (Malme et al. 1985; Madsen et al. 2002). A seismic survey in Puget Sound showed mixed results between species, with some, such as Grey Whales, exhibiting ambiguous responses to the survey while others, such as Harbour Porpoises, tolerating only relatively low exposure levels before leaving the area (Calambokidis et al. 1998).

For obvious ethical reasons, there are no experimental studies of the physical effects of seismic surveys on cetaceans. However the internal structure of the cetacean ear resembles that of both fish and terrestrial mammals (Fay and Popper 2000). A small (20 cu in) airgun has been shown to cause permanent hearing loss in caged fish (McCauley et al. 2003), so it is possible that airguns may be capable of damaging cetacean ears if the whales cannot avoid the sound source. Since Killer Whales are known to be exquisitely dependent on sound for orientation, navigation, locating and catching food, communication, and social interactions, the consequences of severe hearing loss could be fatal.

230 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Commercial sonar

Commercial sonar systems are used in a wide variety of vessels for fishing, navigation (depth sounders), bottom-mapping and detecting obstacles (e.g. side scan sonars). They are generally standard equipment on any vessel over 5 m. These sonars typically generate narrowband sounds at higher frequencies and lower power than military sonars. High frequency sounds are more easily focused into narrow beams and attenuate more quickly than low frequency sounds. Thus the volume of water they influence is smaller. There are many models of commercial sonars, but it is only the units that operate below 100 kHz, the upper limit of Killer Whale hearing, that are of concern. Whales may be able to avoid these sources of sound when boats are widely dispersed, but when boats are concentrated in high traffic areas Killer Whales may have no choice but to travel through heavily ensonified areas.

Shipping

Commercial shipping has increased dramatically in recent years. For example, between 1995 and 1999 the worldwide commercial shipping fleet increased 12% (NRC 2003). There are few studies that have measured changes in the background underwater noise levels over time, but those that have suggest that increased vessel traffic is responsible for the increase in ambient noise over the last 100 years (e.g. Andrew et al. 2002). In the northern hemisphere, shipping noise is the dominant source of ambient noise between 10 to 200 Hz (NRC 2003). While shipping energy is concentrated at low frequencies, ships produce significant amounts of high frequency noise as well. Studies are currently underway to understand and mitigate the impacts of shipping activities and commercial vessel traffic noise on some marine mammal species at risk. For example, in 2017, the Enhancing Cetacean Habitat and Observation (ECHO) initiative led by the Vancouver Fraser Port Authority coordinated a vessel slowdown trial in Haro Strait to better understand and measure the level of noise reduction that can be achieved through reduced vessel speed.

Permitted close approaches

Certain activities have the potential to disturb and/or injure whales because they require physical contact with whales or close approaches by boats for extended periods of time. As a result, in both Canada and the U.S., researchers and filmmakers must obtain federal permits if their projects require close approaches or physical contact with Killer Whales. Close approaches can disturb whales both physically and acoustically. Much of the research on Killer Whales is conducted using boats ranging in size from a few metres to vessels over 30 m, although some is land-based (e.g. Orcalab on Hanson Island, the Robson Bight Marine Warden Program on West Cracroft Island, Johnstone Strait). A boat at 10 m from a whale will be approximately 20 dB louder than a boat at 100 m based on spherical spreading (Richardson et al. 1995).

Photo-identification studies require that all whales in the group be photographed before the encounter is considered complete, and good quality photographs typically mean that whales must be approached to within 30 m (approximately 10 dB louder than at 100 m). Prey fragment sampling, which is providing insight into the diet of Resident Killer Whales, involves approaching the area where a whale has surfaced after it has finished actively feeding. Biopsy darting, a method used in genetic and contaminant studies, also involves close approaches by boats, and recommendations arising from the NOAA Cetacean Systematics Workshop in La Jolla

231 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

California, in April-May 2004 include darting juveniles (Waples and Clapham 2004). The possible health risks of darting young calves have not been evaluated. Some satellite tags and time-depth recorders (TDRs) are applied externally to Killer Whales. They are used to monitor the movements of whales, but may disturb them during the initial application and /or during the time that they adhere to the skin. Newer technologies involving satellite tags and TDRs that are implanted in the skin or muscle pose the additional risk of injuring Killer Whales. From 2013 to 2016, NOAA used satellite tags that attach to Killer Whales through darts that implant into the skin and tissue of these whales. The tags provided information about movements and habitat use of Southern Resident Killer Whales and were used to address knowledge gaps regarding winter distribution of this population. This tagging effort was suspended in April 2016 following the death of Southern Resident L95, after the cause of death was determined to be an infection that was likely introduced through the tag wound (NOAA 2016).

Other forms of disturbance

The number of boats on the water has increased dramatically in recent years. This increase in traffic has the potential to disrupt Killer Whales simply because more vessels are passing through their habitat and potentially disturbing how whales move through the available space. This is most evident when whales are interrupted from their normal activities in order to avoid a collision. While collisions between whales and vessels are relatively rare, when they do occur they can cause significant injury or death (Ford et al. 2000). A science based review of the effectiveness of recovery measures for Southern Resident Killer Whales conducted in 2017 identified vessel strikes as an emerging threat to this population and measures have been suggested to address this threat. Refer to the Review of the Effectiveness of Recovery Measures for Southern Resident Killer Whales (DFO 2017c) for further information.

Personal watercraft (PWC) or ‘jet skis’ may be another potential source of disturbance or injury to Killer Whales. PWC are capable of much more erratic or unpredictable manoeuvres than traditional high speed vessels. As a result they pose a collision risk to Killer Whales and other wildlife. PWC have been banned in the San Juan Islands and in portions of the Monterey Bay National Marine Sanctuary, but they are not banned in the coastal waters of BC, with the exception of the inner waters of Vancouver Harbour. Underwater noise emissions of PWC are reported to consist of broadband energy between 100 Hz and 10 kHz (Erbe 2013).

While Resident Killer Whales must travel in high vessel traffic areas such as Johnstone Strait and the Strait of Georgia, they also must negotiate both commercial and recreational sports fishing boats specifically targeting salmon in ‘hot spots’ that are also good feeding areas for Killer Whales. This includes areas in the vicinity of sports fishing lodges. Conflict for space may force Killer Whales to alter their foraging behaviour in order to successfully capture prey or to avoid collision or entanglement (see section 4.2.5).

Certain industrial activities such as construction, drilling, pile driving, pipe laying and dredging may also disrupt Killer Whales. Construction is also a source of underwater noise. Physical structures, including net pens for aquaculture and permanent structures (e.g. wharves), may damage foraging habitat such as kelp beds, or physically displace Resident Killer Whales from areas they have historically travelled in. If the finfish aquaculture industry continues to expand on the north coast, the placement of net pens may become an issue for Northern Residents.

232 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

4.2.4. Oil spills

While the probability of either Northern or Southern Resident Killer Whales being exposed to a major oil spill is relatively low, the impact of such an event is potentially catastrophic. Both populations are at risk of exposure to an oil spill because of the large volume of tanker traffic that travels in and out of Puget Sound and the Strait of Georgia (Baird 2001; Grant and Ross 2002) and the proposed expansion of tanker traffic along the coast of BC. In 2003, 746 tankers and barges transported over 55 billion litres of oil and fuel through the Puget Sound (WDOE 2004). Though the moratorium on offshore oil and gas exploration and development remains in place in BC, if the moratorium is lifted, the extraction and transport of oil may put Resident Killer Whales at additional risk.

Killer Whales do not appear to avoid oil, as evidenced by the 1989 Exxon Valdez oil spill in Prince William Sound, Alaska. Less than a week after the spill, Resident Killer Whales from one pod were observed surfacing directly in the slick (Matkin et al. 1999). Seven whales from the pod were missing at this time, and within a year, 13 of them were dead. This rate of mortality was unprecedented, and there was strong spatial and temporal correlation between the spill and the deaths (Dahlheim and Matkin 1994; Matkin et al. 1999). The whales probably died from the inhalation of petroleum vapours (Matkin et al. 1999). Exposure to hydrocarbons can be through inhalation or ingestion and has been reported to cause behavioural changes, inflammation of mucous membranes, lung congestion, pneumonia, liver disorders, and neurological damage (Geraci and St. Aubin 1982).

4.2.5. Incidental mortality in fisheries

Killer Whales are rarely entangled in fishing gear, based on anecdotal accounts and an absence of net marks in identification photographs, but the actual numbers of whales caught are unknown (Baird 2001). Several stranded Killer Whales have been found with gear from commercial or recreational line fisheries in their stomachs and the possibility of mortality as a result is unknown (Ford et al. 1998). A few entanglements have been reported from BC, Alaska and California, but they usually have not resulted in death (Pike and MacAskie 1969; Barlow et al. 1994; Heyning et al. 1994; Guenther et al. 1995). In 2014, Northern Resident Killer Whale I103 became severely entangled in a gill net and despite being released quickly, died the following winter. It is likely that entanglement in fishing gear poses little direct threat to Killer Whale populations at present. However, there are areas in BC where Killer Whales have learned to take fish from fishing gear and once this behaviour is adopted, it can spread quickly throughout a population. This problem, referred to as depredation, is severe in many parts of the world (Donogue et al. 2002). Where depredation occurs, deterrent methods, entanglement and accidental hooking can increase the injury or mortality rates of whales.

5. Knowledge gaps

While Resident Killer Whales are among the best studied cetaceans in the world, significant gaps in knowledge about these populations remain. In part this is due to the fact that although studies of Killer Whales have been ongoing over the last 45 years, their whereabouts are poorly known during much of the year. As well, opportunities to learn from Killer Whale carcasses occur relatively infrequently. Only seven to eight carcasses are recovered around the world each year (Raverty et al. 2014). In a 30-year period, only 14 Resident carcasses were found and necropsied in BC (Ford et al. 1998), a recovery rate of 6%.

233 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Some key areas where further knowledge is needed include:

 the year-round distribution and behaviour of Resident Killer Whales  whether potential additional critical habitat areas are required for Resident Killer Whales  the historical abundance of Resident Killer Whales  the year-round diet and energetic requirements of Resident Killer Whales  the consequences of changes in key prey populations on Resident Killer Whales, as well as their historic trends  the population level consequences of low population size and its effects on the sustainability and viability of Resident Killer Whales  the population size that is needed to maintain the cultural and genetic diversity of Resident Killer Whales  the long- and short-term effects of physical disturbance (shipping, whale watching, aircraft, researchers and film makers) on Resident Killer Whales  the long- and short-term effects of acoustic disturbance (whale watching, seismic surveys, military sonar, researchers and film makers) on Resident Killer Whales  the full range of anthropogenic environmental contaminants to which Killer Whales and their prey are exposed, over time and in space, with special attention paid to the identification of sources and the resulting effects of environmental contaminants on Resident Killer Whales, their prey and their habitat  diseases, pathogens, parasites and pathologies of Resident Killer Whales  the effects of climate or environmental change on Resident Killer Whale prey and their habitat

6. Recovery

6.1 Recovery goal

The recovery goal for Northern and Southern Resident Killer Whales is to: ensure the long-term viability of Resident Killer Whale populations by achieving and maintaining demographic conditions that preserve their reproductive potential, genetic variation, and cultural continuity.

The recovery goal reflects, to the extent possible, the complex social and mating behaviour of Resident Killer Whales and the key threats that may be responsible for their decline. In the absence of historical data, it does not identify a numerical target for recovery because our current understanding of Killer Whale population demographics is not adequate for setting a meaningful value at this time. However, because maintaining the demographic conditions that will preserve their reproductive potential, genetic variation, and cultural continuity is fundamental to these populations recovering, a number of demographic indicators are expressed herein that will serve as short-term measures of recovery success. The setting of a quantitative recovery goal will be revisited as new information becomes available.

Killer Whales are top-level predators, and as such will always be far less abundant than most other species in their environment. In addition, they are segregated into small populations that are closed to immigration and emigration, such as the Northern and Southern Resident communities. Furthermore, their capacity for population growth is limited by a suite of life history and social factors, including late onset of sexual maturity, small numbers of reproductive females and mature males, long calving intervals, and dependence on the cultural transmission of ecological and social information. Unfortunately, little is known concerning the historic sizes of

234 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Killer Whale populations, or the factors that ultimately regulate them. Genetic diversity is known to be low in both populations, particularly the Southern Residents, but the consequences of this lack of diversity have not been examined.

6.1.1. Measures of recovery success

The following have been identified as measures of recovery success:

a) long-term maintenance of a steady or increasing size for populations currently at known historic maximum levels and an increasing size for populations currently below known historic maximum levels b) maintenance of sufficient numbers of females in the population to ensure that their combined reproductive potential is at replacement levels for populations at known historic maximum levels and above replacement levels for populations below known historic maximum levels c) maintenance of sufficient numbers of males in the population to ensure that breeding females have access to multiple potential mates outside of their own and closely related matrilines d) maintenance of matrilines comprised of multiple generations to ensure continuity in the transmission of cultural information affecting survival

6.1.2. Monitoring and research strategies

The following monitoring and research programs are essential to define and evaluate the success of the indicators of recovery and will be vital to the establishment of a quantitative recovery goal:

a) routinely monitor Resident Killer Whale population numbers, sex and age composition, social structure and genetic diversity b) develop models of Resident Killer Whale population dynamics and demographics, including social and genetic structure c) develop a quantitative framework to better understand how key anthropogenic and naturally occurring factors, particularly those identified as threats, affect the dynamics of Resident Killer Whale populations d) undertake studies to identify the role of cultural transmission in the foraging ecology, sociobiology and maintenance of genetic diversity in Resident Killer Whales

Because Killer Whale populations are closed to immigration and emigration, and animals individually identifiable, routine monitoring provides accurate, detailed life history information, which is used to determine trends, and to refine and test population models. These models will lead to a better understanding of achievable targets for population recovery. A better understanding of the anthropogenic and naturally occurring factors that regulate or limit Killer Whale populations, and of the role and importance of culture, will make it possible to rank threat factors and prioritize recovery actions.

6.2 Recovery objectives and strategies to achieve recovery

Given our current knowledge, the primary anthropogenic threats to the long-term survival of Northern and Southern Resident Killer Whales appear to be 1) reduced prey availability, 2) environmental contaminants, 3) disturbance, and 4) degradation of critical habitat. We have

235 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 identified four objectives that directly address these threats and contribute to achieving the recovery goal of population viability and sustaining genetic diversity and maintaining cultural continuity (as stated above). The numerical values do not reflect any priority among the objectives. These objectives provide direction for the broad strategies that can be used to specifically mitigate and/or eliminate each of the threats facing Resident Killer Whales, and to better address gaps in our knowledge.

6.2.1 Objective 1

Ensure that Resident Killer Whales have an adequate and accessible food supply to allow recovery.

This objective identifies the need to learn more about the year-round diet of Killer Whales, and to understand and mitigate the threats to key prey populations and their habitat. Food supply can limit the growth and recovery of any population, and there are concerns about the quality and quantity of Resident Killer Whale prey, as well as the prey’s habitat. In some areas of the U.S., for example, runs of Chinook Salmon, a principal prey species for Resident Killer Whales, have been listed as either Endangered or Threatened (NOAA 2017). We know very little about what Killer Whales eat during the winter and spring, and this information is critical to understanding whether the quantity or quality of their food supply could be responsible for the recent decline in Killer Whale numbers, and may prevent their populations from recovering.

Objective 1 strategies

 Determine the seasonal and annual diet and energetic requirements of Resident Killer Whales  Identify key prey populations and feeding areas for Resident Killer Whales  Establish long-term monitoring programs capable of detecting changes in the abundance, distribution and quality of Resident Killer Whale prey  Protect the access of Resident Killer Whales to important feeding areas  Ensure that Resident Killer Whale prey populations and their (the prey’s) habitat are adequately protected from anthropogenic factors such as exploitation and degradation, including contamination, which will allow for the recovery of Resident Killer Whales

6.2.2 Objective 2

Ensure that chemical and biological pollutants do not prevent the recovery of Resident Killer Whale populations.

Ross et al. (2000) showed that Southern Resident Killer Whales are among the most contaminated mammals known, and that Northern Residents also carry significant pollutant loads. These pollutants are known to impair both immune responses and reproduction in other species at lower concentrations than currently seen in Killer Whales. The strategies listed below are intended to improve our understanding of, and mitigate, the contaminant risks that Resident Killer Whales and their prey are exposed to. They also acknowledge the serious risks that pathogens, introduced species and catastrophic events such as oil spills present to Killer Whales and their prey.

236 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Objective 2 strategies

 Investigate the effects of chemical and biological pollutants on the health and reproductive capacity of Resident Killer Whales  Monitor chemical and biological pollutant levels in Resident Killer Whales and their prey  Identify (and prioritize) key chemical and biological contaminants and their sources  Reduce the introduction into the environment of pesticides and other chemical compounds that have the potential to adversely affect the health of Killer Whales and/or their prey, through measures such as national and international agreements, education, regulation, and enforcement  Mitigate the impacts of currently and historically used ‘legacy’ pollutants in the environment  Investigate diseases, pathogens, parasites, and pathologies of Killer Whales  Reduce the introduction of biological pollutants, including pathogens and exotic species, into the habitats of Killer Whales and their prey

In order for these strategies to be successful, it is important that contaminant levels be measured, so as to provide a baseline that can be used to monitor changes in contaminant profiles over time, and to quantify whether attempts at mitigation are successful. Mitigation must occur on scales that range from the local consumer to the international level, as many pollutants originate from sources outside of Canada. Regulations, guidelines and best practices for the manufacture, storage, transport, use and disposal of hazardous compounds must be followed, and evolve to reflect changing knowledge of contaminants and their adverse health effects on Resident Killer Whales, their prey and their habitat. Education at individual, corporate and government levels (again ranging from local to international) will play an important role in reducing the rate at which contaminants are introduced into the environment. International treaties, similar to the Stockholm Convention on Persistent Organic Pollutants, should be endorsed.

6.2.3 Objective 3

Ensure that disturbance from human activities does not prevent the recovery of Resident Killer Whales.

Both physical and acoustic disturbance from human activities may be key factors causing depletion or preventing recovery of Resident Killer Whale populations. Sources of acoustic disturbance range from high-intensity sound produced by seismic surveys to chronic sources such as vessel traffic. During periods of high boating activity in the summer months, disturbance may occur from vessel congestion, impairing the ability of whales to move freely and/ or forage effectively. Physical disturbance can be caused by boat or air traffic close to whales, especially during certain behavioural states such as feeding or beach rubbing (Williams 1999). Research to date has identified various immediate responses of whales to disturbance; however, we know little about potential long-term effects on whale behaviour, health, and foraging efficiency. The National Research Council (NRC 2005) put forward a detailed listing of approaches to better understand how noise impacts marine mammals. The strategies listed here generally address the need for more knowledge about how noise and physical disturbance affect Resident Killer Whales and also provide for mitigation of disturbance as a precautionary measure.

237 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Objective 3 strategies

 Determine the short and long-term effects of chronic and immediate forms of disturbance, including vessels and noise, on the physiology, foraging, and social behaviour of Resident Killer Whales  Determine baseline ambient and anthropogenic noise profiles and monitor sources and changes in the exposure of Resident Killer Whales to underwater noise  Develop and implement regulations, guidelines, sanctuaries and other measures to reduce or eliminate physical and acoustic disturbance of Resident Killer Whales  Develop protocols, regulations, guidelines and/or other measures for the use of underwater seismic survey tools and high energy sonar testing, as most appropriate and in collaboration with stakeholders, to reduce disturbance or injury to Resident Killer Whales, where such activities are permitted

In order to be effective, these strategies will require education and stewardship activities promoting compliance with best practice guidelines, the protection of sanctuaries, and the enforcement of regulations. New technologies, such as those that reduce noise may also contribute to reductions in disturbance over the long-term. Existing regulations, guidelines, protocols and other measures should be evaluated for their efficacy in protecting Resident Killer Whales, particularly as new information becomes available.

6.2.4 Objective 4

Protect critical habitat for Resident Killer Whales and identify additional potential areas for critical habitat designation and protection.

Four areas, used consistently by Resident Killer Whales, are designated as critical habitat as defined by SARA. One, the trans-boundary waters of Haro Strait and Boundary Pass, is used by Southern Residents year-round. The second, the waters of Johnstone and southeastern Queen Charlotte Straits and their adjoining channels, is used by many of the Northern Residents during the summer and fall. The third, which includes waters off southwestern Vancouver Island, is used by members of both the Southern and Northern Resident Killer Whale populations throughout most of the year. The fourth, waters in western Dixon Entrance, is used by Northern Resident Killer Whales year-round. These areas represent a relatively small proportion of the total range of each population and it is unknown whether additional critical habitat may be required to support Resident Killer Whale recovery objectives. Preliminary data suggest that key areas may exist in other locations and at different times of the year, but are not sufficient to warrant proposing these habitats as critical without further research. The strategies listed here provide measures for the protection of the critical habitats referred to above, as well as direction for the potential identification of additional critical habitat.

Objective 4 strategies

 Develop a year-round comprehensive survey program for Resident Killer Whales  Identify key feeding areas and other critical habitat of Resident Killer Whales throughout the year  Protect the access of Resident Killer Whales to their critical habitat  Protect critical habitat areas through assessment and mitigation of human activities that result in contamination and physical disturbance

238 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

 Ensure that sufficient prey is available to Killer Whales in their critical habitat.  Ensure trans-boundary cooperation in the identification and protection of critical habitat

The first two strategies listed above focus on determining whether additional areas should be proposed for critical habitat designation. The remaining strategies, as well as those in Objectives 1, 2 and 3, will help to preserve and protect designated critical habitat.

6.3 Effects on non-target species

The objectives outlined above are designed to protect Resident Killer Whale prey populations and their habitat from exploitation and degradation including contaminants and noise. The spin- off effects of this are likely to be widespread and will be beneficial to human health as well as to a wide variety of organisms ranging from fish to sea birds, since all are affected by contaminants and exploitation. It is likely this benefit will far exceed the increased mortality of prey species associated with increased Killer Whale numbers.

6.4 Evaluation and the status of strategies for recovery

The following are examples of performance measures that may be used to assess the effectiveness of the objectives and strategies, and to determine whether recovery remains feasible. Detailed recovery measures that provide the best chance of meeting the recovery goal and objectives for Northern and Southern Resident Killer Whales, and timelines for their implementation, were identified during the development of the action plan (DFO 2017a).

Progress towards meeting these objectives and strategies was detailed in the Report on the Progress of Recovery Strategy Implementation for Northern and Southern Resident Killer Whales (Orcinus orca) in Canada for the Period 2009-2014 (DFO 2016b) and will continue to be reported on every five years until the objectives have been achieved or until recovery of the species is no longer feasible.

Table 3. Examples of performance measures that may be used to assess the effectiveness of the broad strategies used to achieve the objectives of the recovery strategy for the Northern and Southern Resident Killer Whales in Canada

Examples of performance measures for broad Broad strategy Status* Objective no. strategies and objectives /threat

Recovery Goal: Monitor population dynamics and Underway Completion of annual censuses Ensure long-term demography population Genetic sampling and analyses completed viability Evaluation of population status to ensure growth Develop population models Underway Models developed that incorporate social and genetic structure and explain population trends Quantitative framework for Proposed Models completed that incorporate threats into understanding effects of population dynamic models threats on population dynamics

239 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Examples of performance measures for broad Broad strategy Status* Objective no. strategies and objectives /threat

Studies to identify role of culture Proposed Peer-reviewed publications on role of culture in in foraging ecology and Killer Whale foraging sociobiology Studies to identify role of culture Underway Biopsy samples collected and analyzed to in maintaining genetic identify paternity diversity 1. Ensure Determine seasonal/annual diet/ Underway Prey fragment samples collected year-round adequate and energetic requirements for multiple years accessible food supply Alternative diet sampling methods tested to confirm diet Winter and spring distribution and diet of Resident Killer Whales identified Identify key prey populations and Underway Complete diet sampling of all members of feeding areas population and during all seasons Prey identified to stock, not just species Monitoring prey populations to Underway Population assessment completed for all detect changes in abundance stocks identified as important prey for or availability Resident Killer Whales Protect access to important Proposed Guidelines developed for human activities in feeding areas important whale feeding areas Protection of prey populations Underway Incorporation of Killer Whale predation into fisheries management plans 2. Chemical and Investigate effects of Underway Peer reviewed publication on contaminants in biological contaminants on health and Resident Killer Whales contaminants reproductive capacity of Killer Whales Develop and apply tests to measure the health of Killer Whales Monitor pollutants, diseases, Underway Extensive sampling of populations to establish pathogens, parasites and baseline contaminant levels pathologies in Killer Whales Completed analyses of contaminants in samples Complete necropsies of stranded Killer Whales. Identify and prioritize key Underway Completed sampling and analyses of chemical and biological contaminants in Killer Whale prey pollutants Identify and prioritize key Underway Water quality sampling in areas throughout sources of chemical and range of Resident Killer Whales biological pollutants

Reduce introduction of chemical Underway Measurable decline in contaminant levels in pollutants into environment environment (prey, sediments etc.) Mitigate impacts of currently Underway Evaluation of effectiveness of legislation used pollutants completed

240 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Examples of performance measures for broad Broad strategy Status* Objective no. strategies and objectives /threat

Mitigate impacts of ‘legacy’ Underway PCB sources identified pollutants Reduce introduction of biological Underway Evaluation of effectiveness of legislation pollutants completed 3. Acoustical and Investigate short-term effects of Underway Controlled studies of whale/boat interactions Physical chronic forms of disturbance completed Disturbance Investigate short-term effects of Proposed Complete controlled study of marine mammals acute forms of disturbance in areas where seismic exploration is active Investigate long-term effects of Proposed Complete model that incorporates effects of chronic forms of disturbance increasing ambient noise levels on communication signals of Resident Killer Whales Investigate long-term effects of Proposed Complete controlled study of marine mammals acute forms of disturbance in areas where seismic exploration is active Determine baseline ambient and Proposed Complete acoustic profiles of vessels most anthropogenic noise profiles likely to be encountered by Resident Killer Whales Develop measures to reduce Underway Revised whale watching guidelines, and/ or physical disturbance regulations that reflect most recent understanding of effects of chronic physical disturbance Develop measures to reduce Proposed Establishment of acoustic sanctuaries in acoustic disturbance critical habitat areas Develop measures for reducing Proposed Revised protocols for seismic and military disturbance to high energy sonar that reflect most recent sources of sound understanding of physiological and behavioural responses to noise 4. Protection of Year-round comprehensive Underway Winter distribution of Resident Killer Whales critical habitat surveys to identify important well understood areas for Killer Whales Identify key feeding areas and Underway Winter prey of Resident Killer Whales other critical habitat identified Protect access of whales to Underway Sanctuaries within critical habitat established critical habitat Protect critical habitat from Proposed Measurable reduction in contaminants in contamination, and physical critical habitat disturbance Ensure sufficient prey available Proposed Key prey populations in critical habitat areas to whales in critical habitat Ensure trans-boundary Proposed Formal identification of critical habitat cooperation in identification recognized by international agreement and protection of critical habitat

241 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Note: A thorough listing of recovery measures was included in the action plan (DFO 2017a).

* The information presented in the status column of this table represents the status of each strategy at the time of the development of the original recovery strategy (2008). See Table 3 in the Report on the Progress of Recovery Strategy Implementation for Northern and Southern Resident Killer Whales (DFO 2016b) for more recent information regarding progress made toward each of these objectives and strategies.

7. Critical habitat

7.1 Identification of the species’ critical habitat

7.1.1 General description of the species’ critical habitat

Critical habitat is defined in SARA section 2(1) as “…the habitat that is necessary for the survival or recovery of a listed wildlife species and that is identified as the species’ critical habitat in a recovery strategy or in an action plan for the species.”

SARA defines habitat for aquatic species at risk as “… spawning grounds and nursery, rearing, food supply, migration and any other areas on which aquatic species depend directly or indirectly in order to carry out their life processes, or areas where aquatic species formerly occurred and have the potential to be reintroduced” [s. 2(1)].

Partial critical habitat was identified for both Northern and Southern Resident Killer Whales in the 2008 recovery strategy. Northern Resident Killer Whale critical habitat included the waters of Johnstone Strait and southeastern Queen Charlotte Strait (Figure 4), while Southern Resident Killer Whale critical habitat included the transboundary waters in southern BC, including the southern Strait of Georgia, Haro Strait, and Juan de Fuca Strait (Figure 5). These critical habitat areas were protected through the making of a SARA Critical Habitat Order in 2009. In 2011, minor amendments were made to the critical habitat section of the 2008 recovery strategy. These amendments clarified that attributes of critical habitat identified in the 2008 recovery strategy are a part of critical habitat.

Two additional areas were identified for consideration as critical habitat for Resident Killer Whales in DFO (2017b).These areas include: i) waters on the continental shelf off southwestern Vancouver Island, including Swiftsure and La Pérouse Banks (Northern and Southern Resident Killer Whale critical habitat, Figures 4 and 5) and ii) the waters of western Dixon Entrance along the north coast of Graham Island from Langara Island to Rose Spit (Northern Resident Killer Whale critical habitat, Figure 4).

For Northern and Southern Resident Killer Whales, critical habitat is identified in this recovery strategy to the extent possible, using the best available information. A description of the functions, features, and attributes that support the identification of critical habitat is provided in section 7.1.3.

Under SARA, critical habitat must be legally protected within 180 days of being identified in a recovery strategy or action plan through a SARA Critical Habitat Order or under any other Act of Parliament, and through prohibitions to the destruction of any part of critical habitat.

242 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

This recovery strategy identifies critical habitat for Resident Killer Whales as four distinct geographic areas. These include: 1) the waters of Johnstone Strait and southeastern Queen Charlotte Strait (Northern Resident Killer Whale critical habitat); 2) transboundary waters in southern British Columbia, including the southern Strait of Georgia, Haro Strait, and Juan de Fuca Strait (Southern Resident Killer Whale critical habitat); 3) waters on the continental shelf off southwestern Vancouver Island, including Swiftsure and La Pérouse Banks (Northern and Southern Resident Killer Whale critical habitat); and 4) waters of west Dixon Entrance, along the north coast of Graham Island from Langara to Rose Spit (Northern Resident Killer Whale critical habitat).

It is unknown if the critical habitat identified in this recovery strategy is sufficient to achieve the species’ recovery goal and objectives. The schedule of studies outlines the research required to identify additional critical habitat and to acquire more detail about the critical habitat identified to achieve the species’ recovery goal and objectives. Additional critical habitat may be identified in future amendments to this recovery strategy.

7.1.2 Information and methods used to identify critical habitat

The movement patterns of Resident Killer Whales are influenced by the availability of their preferred prey. During the summer and fall months, Resident Killer Whale distribution is associated spatially and temporally with the migratory routes of Chinook Salmon as this important prey species returns to natal streams to spawn (Ford and Ellis 2005). For the rest of the year there is less information available on the diet, distribution, and movement patterns of Resident Killer Whales, though surveys, passive acoustic monitoring, and satellite tagging studies have been conducted to address these knowledge gaps (Riera 2012; Hanson et al. 2013; DFO 2017b). Determining whether there are additional habitats that the whales utilize during winter and spring that are critical for the survival or recovery of these populations has been and continues to be a priority. This will need to take into account the likelihood that changes in the availability of major stocks of key prey species may cause corresponding shifts in the geographic location of critical habitat for Resident Killer Whales.

Methods and rationale used to designate each of the four critical habitat areas identified to date are detailed below.

Northern Resident Killer Whale critical habitat: Johnstone Strait and southeastern Queen Charlotte Strait

Analyses of existing data on coast-wide occurrence patterns of Northern Resident Killer Whales provide quantitative documentation of the importance of Johnstone Strait and southeastern Queen Charlotte Strait (Figure 4) to these whales (Ford 2006). These analyses, along with previously published information, form the basis for this area’s critical habitat designation.

One or more Northern Resident matrilines are sighted in this area on most days during July through October, with peak numbers generally in mid-July to mid-September (Nichol and Shackleton 1996; Ford 2006). Sightings become more sporadic in the area during November, are scarce from December through May. Although all Northern Resident pods have been identified in the area, different pods do not use the area equally (Ford et al. 2017). For example, 75% of encounters documented during 1990-2004 included all or part of A1 pod, while only 0.7% of encounters during this same period included I18 pod, a group of similar size. Northern Resident Killer Whales in the Johnstone Strait area spend the majority of time foraging for

243 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 salmon, primarily Chinook during July-September and Chum in October (Ford 1989; Ford et al. 1998; Ford 2006; Ford et al. 2010; DFO 2017b). Other activities undertaken in the area include resting, socializing, and beach rubbing (Ford 1989; Ford et al. 2000; Ford 2006).

Beach rubbing appears to be an important activity for Northern Resident Killer Whales. More than 90% of the Northern Resident Killer Whales observed in Johnstone Strait visit the rubbing beaches, and spend about 10% of their time there (Briggs 1991). During this time they are very sensitive to disturbance. In recognition of the importance of this habitat to Resident Killer Whales, in 1982 the Province of British Columbia established the Robson Bight–Michael Bigg Ecological Reserve to protect a portion of western Johnstone Strait and the foreshore near Robson Bight, where the rubbing beaches are located. This Ecological Reserve includes the primary foraging areas for Killer Whales utilizing the Johnstone Strait area, as well as at least six beaches used to various degrees by these whales for rubbing, and is included as critical habitat within the Johnstone and southeastern Queen Charlotte Straits boundaries (see Table 4).

Given the importance of this area to a significant component of the Northern Resident community for a major portion of the salmon feeding season, and the traditional use of rubbing beaches located there, this area has been designated as critical habitat as defined in SARA.

Southern Resident Killer Whale critical habitat: transboundary waters of the southern Strait of Georgia, Haro Strait, and Juan de Fuca Strait

The transboundary waters of southern BC and Washington State (Figure 5) represent an important area of high concentration of Southern Resident Killer Whales. This area includes waters under both Canadian and U.S. jurisdiction. Analyses of existing data on coast-wide occurrence patterns of Southern Resident Killer Whales have been completed by NOAA as part of the ESA designation of critical habitat in collaboration with DFO (NMFS 2006a). This assessment provided quantitative documentation of the importance of these transboundary areas to these whales and forms, along with previously published information, the basis for the critical habitat identification.

This critical habitat area is utilized regularly by all three Southern Resident pods during June through October in most years (Osborne 1999; Wiles 2004). J pod appears to be present in the area throughout much of the remainder of the year, but two Southern Resident pods, K and L, are typically absent during December through April. This critical habitat is of great importance to the entire Southern Resident community as a foraging range during the period of salmon migration, and thus has been designated as critical habitat under SARA.

Northern and Southern Resident Killer Whale critical habitat: southwestern Vancouver Island; and Northern Resident Killer Whale critical habitat: western Dixon Entrance

Southwestern Vancouver Island and western Dixon Entrance were identified as habitats of special importance to Northern and Southern Resident Killer Whales based on photo- identification, predation, and acoustic data. Passive acoustic monitoring was used to supplement the boat-based photo-identification studies, as these areas are remote and exposed to open ocean conditions, making small boat studies difficult. Using both photo-identification and detections of Resident Killer Whale vocalizations on acoustic recording devices allowed for year-round assessment of Resident Killer Whale occurrence in these areas. Tissue samples and scales were collected at predation sites to identify prey and assess Resident Killer Whale

244 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 diet. See DFO (2017b) for detailed information about the methods used to identify these two areas as Resident Killer Whale critical habitat.

Critical habitat off southwestern Vancouver Island (Figures 4 and 5) includes the Canadian portions of Swiftsure Bank, where acoustic monitoring between August 2009 and July 2011 indicated considerable habitat use by both Southern and Northern Resident Killer Whales over much of the year. Additionally, it encompasses several other relatively shallow banks, including La Pérouse Bank which, like Swiftsure Bank, is among the most productive fishing areas for Chinook Salmon on the west coast of North America. During this acoustic monitoring, all three Southern Resident Killer Whale pods were detected in this area, with L pod being the most frequently documented (Ford et al. 2017). The area is important for Southern Resident Killer Whales, both during summer, when groups of whales spend time west of the critical habitat area in the transboundary waters in southern BC, and in winter, when whales are mostly absent from the southern BC critical habitat area, but were detected frequently off southwestern Vancouver Island (DFO 2017b). Northern Resident Killer Whales were detected in all months of the year, especially in March and April. Fifteen of the 16 pods in the Northern Resident Killer Whale population were also encountered during boat surveys in this area. Given the importance of southwestern Vancouver Island to both Northern and Southern Resident Killer Whales throughout most of the year, it meets the definition for critical habitat under SARA for both of these populations.

Critical habitat in western Dixon Entrance (Figure 4) is an important foraging area for Northern Resident Killer Whales, and is situated on migratory routes for a wide variety of Chinook Salmon stocks. Northern Resident Killer Whales were detected during acoustic monitoring activities between September 2009 and June 2012 in this area in all months of the year, but most frequently during late winter and spring (DFO 2017b). Certain pods that are rarely encountered in Northern Resident Killer Whale critical habitat in Johnstone and eastern Queen Charlotte Straits at any time of year were frequently detected in western Dixon Entrance (DFO 2017b). As habitat needed for recovery, western Dixon Entrance appears to be far more relevant for these pods than the Johnstone /Queen Charlotte Straits critical habitat area. Pods belonging to G clan were documented in western Dixon Entrance most often, followed by R01 pod (R clan). The use of western Dixon Entrance by members of the population that are rarely documented in the Johnstone Strait area and during times of year when detections of this population in the Johnstone Strait area are infrequent indicates its importance to Northern Resident Killer Whales. This area thus meets the definition for critical habitat under SARA.

7.1.3 Identification of critical habitat

Geographic information

Four critical habitat areas have been identified for Resident Killer Whales, described in section 7.1.1. Critical habitat was geographically delineated and is identified as the areas within the identified geographical boundaries (Appendix D), given that they contain the described biophysical features, attributes, and the functions they support. The critical habitat’s biophysical functions, features, and attributes are described below and summarized in Table 4, and need to be protected against destruction.

245 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Northern Resident Killer Whale critical habitat: Johnstone Strait and southeastern Queen Charlotte Strait

The boundaries of the existing critical habitat area for Northern Resident Killer Whales include the waters of Johnstone Strait and southeastern Queen Charlotte Strait, and the channels connecting these straits as depicted in Figure 4. This area is approximately 905 km2.

Southern Resident Killer Whale critical habitat: Transboundary waters of the southern Strait of Georgia, Haro Strait, and Juan de Fuca Strait

Critical habitat for Southern Resident Killer Whales includes the transboundary areas of southern BC and Washington State. The portion of this critical habitat that is in Canadian waters is approximately 3,390 km2 in size, and includes the Canadian side of Haro and Juan de Fuca Straits, as well as Boundary Pass and adjoining areas in the Strait of Georgia, as depicted in Figure 5.

Much of the area that qualifies as critical habitat for Southern Resident Killer Whales falls within U.S. jurisdiction, and the identification of critical habitat under SARA only applies to the portion of the area that is within Canadian waters (Figure 5). In November 2005, the U.S. listed Southern Resident Killer Whales as Endangered under the ESA (NMFS 2006a). As a result, 6,630 km2 of U.S. inland waters of Washington State and Juan de Fuca Strait were designated as critical habitat under the ESA in November 2006 (NMFS 2006b, see Figure 5).

Northern and Southern Resident Killer Whale critical habitat: Southwestern Vancouver Island

Critical habitat for Northern and Southern Resident Killer Whales located off southwestern Vancouver Island forms a contiguous westward extension of the critical habitat area for Southern Resident Killer Whales described above. The southern boundary is formed by the Exclusive Economic Zone of Canada and extends to the 200 m isobath, or depth contour. See Figures 4 and 5 for the boundaries of this critical habitat, which encompass an area of 5,025 km2.

Northern Resident Killer Whale critical habitat: Western Dixon Entrance

Critical habitat for Northern Resident Killer Whales in western Dixon Entrance includes most of the coastal waters off the north side of Graham Island. The shallow waters of Naden Harbour, Massett Inlet, and McIntyre Bay are not included as critical habitat, due to limited use of these areas by Resident Killer Whales. See Figure 4 for the boundaries of this critical habitat, which encompass an area of 1,394 km2.

246 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Figure 4. Critical habitat areas identified for Northern Resident Killer Whales. Critical habitat is identified as the areas within the identified geographic boundaries, given that they contain the described biophysical features and the functions they support, as described in Table 4.

247 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Figure 5. Critical habitat areas identified for Southern Resident Killer Whales. Critical habitat is identified as the areas within the identified geographic boundaries, given that they contain the described biophysical features and the functions they support, as described in Table 4. The hatched areas in the transboundary waters of southern BC and off southwestern Vancouver Island are the critical habitat areas in Canadian waters for Southern Resident Killer Whales, as designated under SARA. The hatched area in the transboundary waters of northern Washington State is designated as Southern Resident Killer Whale critical habitat under the U.S. ESA.

Biophysical functions, features, and attributes

Seasonal distribution and movement patterns of Resident Killer Whales in Canadian Pacific waters are strongly associated with the availability of their preferred prey, Chinook Salmon, and secondarily Chum Salmon (Ford 2006; Ford and Ellis 2006; Ford et al. 2010; Hanson et al. 2010). Habitats that are important for the survival or recovery of Resident Killer Whales are those that provide for profitable foraging on these key prey species, including the acoustic and physical space required to successfully pursue and capture prey. Other activities, including

248 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 resting and socializing, similarly depend on an acoustic environment that does not impede effective communication among whales.

The only activity that is strictly associated with particular geographic locations is beach rubbing by Northern Resident Killer Whales, which only takes place at specific traditional sites. Several of these sites are included in the Northern Resident Killer Whale critical habitat located in Johnstone Strait (Ford 2006). These rubbing beaches are composed of small rounded pebbles approximately 1 – 5 cm in diameter and are usually situated along otherwise rocky shorelines (Ford pers. comm. 2018).

Table 4 summarizes the best available knowledge of the biophysical functions, features, and attributes for Northern and Southern Resident Killer Whales. The features, functions, and attributes described in Table 4 are based on narrative provided in the 2011 recovery strategy and on Table 1 of DFO (2017b). They apply to all four critical habitat areas and to both populations, with the exception of those associated with beach rubbing, a function that is known to be important for Northern Resident Killer Whales but has not been documented for Southern Residents. As Resident Killer Whales travel with their matrilines throughout their lives, the features, functions, and attributes described also apply to all life stages of Resident Killer Whales.

There is currently insufficient information with which to quantify the levels of many of the attributes listed in Table 4 required to support the features and functions of critical habitat. For example, the density, quantity, and quality of prey needed to support Resident Killer Whale populations are unknown. Additionally, although it is assumed that Chinook remains the primary prey species of Resident Killer Whales throughout the year, the vast majority of Resident Killer Whale prey samples have been collected during summer and fall, and their year-round diet is not well-understood. It is therefore possible that additional important prey species may be identified in the future. Broad studies focused on identifying additional habitats that are important to Resident Killer Whales, and to better understand threats to critical habitat are included in section 7.2. In addition, the action plan for Northern and Southern Resident Killer Whales includes several recovery measures to refine knowledge of the functions, features and attributes of critical habitat. The descriptions of the attributes in Table 4 may be refined in the future, as additional information becomes available.

Table 4. Summary of the biophysical functions, features, and attributes of critical habitat necessary for the survival or recovery of Northern and Southern Resident Killer Whales Function Feature Attribute Feeding and foraging Availability of Chinook Sufficient quantity and quality of Salmon, Chum Salmon, and Chinook Salmon to provide for other important prey species profitable foraging

Diversity of Chinook stocks with a variety of spatial and temporal migration patterns sufficient to maintain availability

Sufficient quantity and quality of Chum Salmon and other species that comprise part of the Resident Killer Whale diet

249 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Feeding and foraging Acoustic environment Anthropogenic noise level that does not interfere with life functions and is Reproduction, socializing, sufficient for effective acoustic social resting signaling and echolocation to locate prey Beach rubbing (Northern Resident Killer Whales) Anthropogenic noise level that does not result in loss of habitat availability or function Feeding and foraging Water quality Water quality of a sufficient level to support Chinook stocks Reproduction, socializing, resting Water quality of a sufficient level to support Chum Salmon and other species that comprise part of the Resident Killer Whale diet

Water quality of a sufficient level so as not to result in loss of function Feeding and foraging Physical space Unimpeded physical space surrounding individual whales Reproduction, socializing, (minimum vessel approach distance resting 200m) Beach rubbing (Northern Rubbing beach Suitable physical habitat to allow for Resident Killer Whales) beach rubbing behaviour

Summary of critical habitat relative to the recovery goal and objectives

Critical habitat areas identified in this recovery strategy are areas that, based on current best available information, the Minister of Fisheries and Oceans and the Minister responsible for the Parks Canada Agency consider necessary to partially achieve the recovery goal and objectives required for the survival or recovery of Northern and Southern Resident Killer Whales.

It is unknown if critical habitat identified in this recovery strategy is sufficient to achieve the species’ recovery goal and objectives. The schedule of studies outlines the research required to identify additional critical habitat and to acquire more detail about the critical habitat identified to achieve the species’ recovery goal and objectives. Additional critical habitat may be identified in future updates to the recovery strategy.

7.2 Schedule of studies to identify critical habitat

Further research is required to refine the understanding of the functions, features, and attributes of the currently identified critical habitat, to identify potential additional critical habitat necessary to support the species’ recovery goal and objectives and to protect the critical habitat from destruction. This additional work includes the studies listed in Table 5. Refer to the report on the progress of recovery strategy implementation for the Northern and Southern Resident Killer Whales in Canada (DFO 2016b) for more details regarding the progress made and the status of the studies outlined in Table 5. In addition to the broad studies outlined in Table 5, the Action Plan for the Northern and Southern Resident Killer Whales (Orcinus orca) in Canada (DFO 2017a) includes more specific recovery measures focused on refining the understanding of the

250 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 features, functions, and attributes outlined in Table 4, on threats to critical habitat and on supporting identification of additional areas for critical habitat identification.

Table 5. Schedule of studies to refine critical habitat and identify potential additional critical habitat areas for Northern and Southern Resident Killer Whales Study Status

Year-round comprehensive surveys to identify areas of occupancy Underway

Identify key feeding areas throughout the year to determine whether they Underway should be proposed as additional critical habitat

Identify activities other than foraging that may be important functions of Underway critical habitat

Identify sources of acoustic disturbance that may negatively impact or Underway affect access to critical habitat

Identify sources of physical disturbance that may negatively impact or Underway affect access to critical habitat

Identify sources of biological and chemical contaminants that may Underway negatively impact critical habitat

Identify factors that may negatively affect an adequate and accessible Underway supply of prey in areas of critical habitat

7.3 Activities likely to result in the destruction of critical habitat

Under SARA, critical habitat must be legally protected within 180 days of being identified in a final recovery strategy or action plan and included in the Species at Risk Public Registry. Northern Resident Killer Whale critical habitat in Johnstone and southeastern Queen Charlotte Straits and Southern Resident Killer Whale critical habitat in the transboundary waters of southern BC has been protected through a SARA Critical Habitat Order since 2009. It is anticipated that protection of the four Resident Killer Whale critical habitat areas identified in this recovery strategy will be accomplished through SARA Critical Habitat Orders made under subsections 58(4) and (5), which will invoke the prohibition in subsection 58(1) against the destruction of any part of the identified critical habitat. The term “recovery strategy” is defined in subsection 2(1) of SARA as meaning “a recovery strategy included in the public registry under subsection 43(2), and includes any amendment to it included in the public registry under section 45”. Thus, reading the definition of “critical habitat” in conjunction with the definition of “recovery strategy,” the Critical Habitat Order will apply not only to the critical habitat identified in the recovery strategy, but also to any modification to the critical habitat subsequently made in an amended recovery strategy, without the need to amend the Critical Habitat Order.

The following examples of activities likely to result in the destruction of critical habitat are based on known human activities that are likely to occur in and around critical habitat and would result in the destruction of critical habitat if unmitigated. Some activities may impact critical habitat regardless of whether or not the whales are present within the area, while others would require the presence of the whales, dependent on the activity and the feature, function, or attribute affected by that activity. The list of activities is neither exhaustive nor exclusive and has been

251 Recovery Strategy for Northern and Southern Resident Killer Whales 2018 guided by the threats described in section 4 of this recovery strategy. The absence of a specific human activity does not preclude or restrict the Department or Parks Canada Agency’s ability to regulate that activity under SARA. Furthermore, the inclusion of an activity does not result in its automatic prohibition and does not mean the activity will inevitably result in destruction of critical habitat. Every proposed activity must be assessed on a case-by-case basis and site-specific mitigation will be applied where it is reliable and available. Where information is available, thresholds and limits have been developed for critical habitat attributes to better inform management and regulatory decision making. However, in many cases knowledge of a species and its critical habitat’s thresholds of tolerance to disturbance from human activities is incomplete.

Reduced prey availability

Availability of Chinook and Chum Salmon is key to the presence of Resident Killer Whales in critical habitat areas (Ford and Ellis 2005; DFO 2017b). Maintaining an adequate food supply for Resident Killer Whales depends on many factors, including the body size and condition of prey, as well as sufficient stock diversity of key prey species to maintain prey availability over time. Activities that result in insufficient abundance, quality, or availability of Chinook Salmon, Chum Salmon, or other Resident Killer Whale prey species could therefore lead to destruction of critical habitat. These include fishing for Chinook Salmon, Chum Salmon, and other important prey species, as well as activities that impact the survival and prey supply of these species such that they are not of sufficient abundance, quality, or availability for Resident Killer Whales.

Acoustic disturbance

There is growing awareness of the impacts of anthropogenic noise on cetaceans (Nowacek et al. 2007; Weilgart 2007). As Killer Whales rely on sound to carry out their life functions, including foraging and socializing, the acoustic environment is an important component of critical habitat. Threats to the acoustical integrity of critical habitat include both acute and chronic noise, and are discussed in detail in section 4.2.3 Underwater noise. Acute noise, including seismic surveys, military and commercial sonars, pile driving and underwater explosions, can result in behavioural changes and displacement from habitat for cetaceans (Morton and Symonds 2002; Weilgart 2007). Chronic noise is primarily associated with vessel traffic, and can result in masking of communication and echolocation signals of Killer Whales (Erbe 2002; Foote et al. 2004; Holt et al. 2009).

Environmental contaminants

Environmental contaminants pose a serious threat to Killer Whales. These contaminants and their sources are discussed in section 4.2.1. As high trophic level, long-lived animals, Killer Whales are particularly vulnerable to persistent bioaccumulating toxins (PBTs) that accumulate in their fatty tissues as they feed on already contaminated prey. The introduction of high levels of contaminants is therefore a threat to Resident Killer Whale critical habitat. While many contaminants are airborne and dispersed throughout the coastal waters of BC, the waters surrounding the lower mainland and Vancouver Island are particularly at risk due to their proximity to human settlement.

The threat of a spill of oil or other toxic material within the areas of critical habitat poses not only an immediate and acute risk to the health of Resident Killer Whale populations (see section 4.2.4), but has the potential to make critical habitat areas uninhabitable for an extended period of time.

252 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Physical disturbance

Vessels in close proximity to Killer Whales have the potential to disrupt behaviours, including foraging and beach rubbing (Williams et al. 2006; Lusseau et al. 2009). Additionally, prey must be physically accessible to Resident Killer Whales and physical obstacles at the surface and underwater which interfere with whales’ abilities to pursue and capture prey represent a threat to critical habitat.

Killer Whales are particularly sensitive to disturbance while beach rubbing (Williams et al. 2006); the physical presence of vessels and other obstacles can not only disrupt beach rubbing, but can also prevent Killer Whales from approaching rubbing beaches to initiate this behaviour.

Geophysical disturbance

A key feature of the Northern Resident Killer Whale critical habitat in Johnstone and southeastern Queen Charlotte Straits is the presence of several rubbing beaches. Activities associated with the geophysical destruction of rubbing beaches are therefore threats to critical habitat. Rubbing beaches may also be vulnerable to activities that alter stream flow and siltation; thus, activities that occur upstream of these beaches, even outside of the designated critical habitat area, can pose a threat to critical habitat.

253 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Table 6. Examples of activities likely to result in the destruction of critical habitat of Northern and Southern Resident Killer Whales Threat Activity Effect - pathway Function affected Feature affected Attribute affected Reduced prey Fishing for Chinook Loss of prey Feeding and foraging Availability of Chinook Sufficient quantity and availability Salmon, Chum Salmon, Salmon, Chum Salmon quality of Chinook and other important Loss of forage fish for and other important Salmon to provide for prey species prey species prey species profitable foraging

Other activities that are Diversity of Chinook detrimental to habitat stocks with a variety of and survival of prey spatial and temporal (e.g. fishing for forage migration patterns fish species) sufficient to maintain availability

Sufficient quantity and quality of Chum Salmon and other species that comprise part of the Resident Killer Whale diet Acoustic disturbance Vessel traffic Chronic noise Feeding and foraging Acoustic environment Anthropogenic noise resulting in masking of level that does not Seismic surveys, communication and Reproduction, interfere with life military, and commercial echolocation socializing, resting functions and is sonars sufficient for effective Acute and chronic Beach rubbing acoustic social signaling Pile driving, underwater acoustic disturbance (Northern Resident and echolocation to explosions resulting in disruption Killer Whales) locate prey of behaviour or displacement from Anthropogenic noise habitat level that does not result in loss of habitat availability or function Environmental Release of deleterious Loss of prey or Feeding and foraging Availability of Chinook Water quality of a contaminants substances reduction in prey Salmon, Chum Salmon, sufficient level to quality Reproduction, and other important support Chinook stocks Point source and non- socializing, resting prey species point source pollution Loss of water quality Water quality of a Water quality sufficient level to support Chum Salmon and other species that comprise part of the Resident Killer Whale diet

254 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Water quality of a sufficient level so as not to result in loss of function Physical disturbance Vessel approach to Reduction of physical Feeding and foraging Physical space Unimpeded physical whales (within 200m) space available to space surrounding whales Reproduction, individual whales Vessel anchoring in socializing, resting (minimum vessel vicinity of rubbing Displacement of approach distance beaches whales from rubbing Beach rubbing 200m) beaches, prevention (Northern Resident Activities that prevent of use of rubbing Killer Whales) approach of whales to beaches rubbing beaches; or that displace or disrupt rubbing behaviour

Human presence on rubbing beaches when whales present or nearby Geophysical Shore-based industrial Geophysical Beach rubbing Rubbing beach Suitable physical habitat disturbance activities that could alter disturbance resulting (Northern Resident to allow for beach beach substrate in loss of function Killer Whales) rubbing behaviour

Activities that result in alteration of stream flow to rubbing beaches, beach sediment, and siltation

255 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

8. Statement on action plans

SARA’s approach to recovery planning is a two-part approach, the first part being the recovery strategy and the second part being the action plan. An action plan contains specific recovery measures or activities required to meet the objectives outlined in the recovery strategy.

An action plan for Northern and Southern Resident Killer Whales was completed and posted to the Species at Risk Public Registry in 2017 (DFO 2017a). Additionally, several multi-species action plans that include Resident Killer Whales have been developed by the Parks Canada Agency and posted to the Species at Risk Public Registry. These include the multi-species action plan for Gwaii Haanas National Park Reserve, National Marine Conservation Area Reserve, and Haida Heritage Site (Parks Canada Agency 2016); the multi-species action plan for Pacific Rim National Park Reserve of Canada (Parks Canada Agency 2017); and the multi- species action plan for Gulf Islands National Park Reserve of Canada (Parks Canada Agency 2018).

256 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

9. References Addison, R.F. and W.T. Stobo. 2001. Trends in organochlorine residue concentrations and burdens in grey seals (Halichoerus grypus) from Sable Is., N.S., Canada, between 1974 and 1994. Environmental Pollution 112: 505-513. Aguilar, A. and A. Borrell. 1994. Abnormally high polychlorinated biphenyl levels in striped dolphins (Stenella coeruleoalba) affected by the 1990-92 Mediterranean epizootic. Science of the Total Environment 154: 237-247. Amos, B., C. Schloetterer, and D. Tautz. 1993. Social structure of pilot whales revealed by analytical DNA profiling. Science 260: 670-672. Andrew, R.K., B.M. Howe, J.A. Mercer, and M.A. Dzieciuch. 2002. Ocean ambient sound: comparing the 1960s with the 1990s for a receiver off the California coast. Acoustic Research Letters Online (ARLO) 3: 65-70. Au, W.W.L., J.K.B. Ford, and K.A. Newman Allman. 2004. Echolocation signals of free-ranging Killer Whales (Orcinus orca) and modeling of foraging for Chinook Salmon (Oncorhynchus tshawytscha). Journal of the Acoustical Society of America 221: 559- 564. Bain, D.E. 1990. Testing the validity of inferences drawn from photo-identification data, with special reference to the studies of the Killer Whales (Orcinus orca) in British Columbia. Report of the International Whaling Commission Special Issue 12: 93-100. Bain, D.E. 1996. Sound level contours produced by the 1995 acoustic barrier at the Hiram M. Chittenden Locks. NMFS Contract Report No. 40ABNF502019. 18pp Bain, D.E. 2002. A model linking energetic effects of whale watching to Killer Whale (Orcinus orca) population dynamics. Available at: https://www.researchgate.net/publication/237227404_ [accessed March 2018]. Bain, D.E. and M.E. Dahlheim. 1994. Effects of masking noise on detection thresholds of killer whales. pp. 243-56. In: T.R. Loughlin (ed.) Marine Mammals and the Exxon Valdez. Academic Press, San Diego. Baird, R.W. 2000. The Killer Whales, foraging specializations and group hunting. Pages 127- 153 in J. Mann, R.C. Connor, P.L. Tyack, and H. Whitehead (editors). Cetacean societies: field studies of dolphins and whales. University of Chicago Press, Chicago, Illinois. Baird, R.W. 2001. Status of Killer Whales, Orcinus orca, in Canada. Canadian Field Naturalist 115:676-701. Baird, R.W. 2002. Killer Whales of the world: natural history and conservation. Voyageur Press, Stillwater, Minnesota. Baker, C.S., G.M. Lento, F. Cipriano, and S.R. Palumbi. 2000. Predicted decline of protected whales based on molecular genetic monitoring of Japanese and Korean markets. Proceedings of the Royal Society of London Series B: Biological Sciences 267: 1191- 1199. Barlow, J., R.W. Baird, J.E. Heyning, K. Wynne, A.M. Manville, L.F. Lowry, D. Hanan, J. Sease, and V.N. Burkanov. 1994. A review of cetacean and pinniped mortality in coastal

257 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

fisheries along the west coast of the USA and Canada and the east coast of the Russian Federation. Report of the International Whaling Commission Special Issue 15:405-425. Barrett-Lennard, L.G. 2000. Population structure and mating patterns of Killer Whales as revealed by DNA analysis. Ph.D. Thesis, University of British Columbia, Vancouver, British Columbia. Barrett-Lennard, L.G. and G.M. Ellis. 2001. Population structure and genetic variability in Northeastern Pacific Killer Whales: toward an assessment of population viability. CSAS Res. Doc. 2001/065. 35 pp. Barrett-Lennard, L.G., J.K.B. Ford and K. Heise. 1996. The mixed blessing of echolocation: Differences in sonar use by fish-eating and mammal-eating Killer Whales. Animal Behaviour 51: 553-565. Beamish, R.J., C. Mahnken, and C.M. Neville. 1997. Hatchery and wild production of Pacific salmon in relation to large-scale, natural shifts in the productivity of the marine environment. ICES Journal of Marine Science 54: 1200-1215 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. Berzin, A.A. and V.L. Vladimorov. 1983. A new species of Killer Whale (Cetacea, Delphinidae) from Antarctic waters. Zoologicheskii Zhurnal 62: 287-295. Bigg, M.A., P.F. Olesiuk, G.M. Ellis, J.K.B. Ford, and K.C. Balcomb. 1990. Social organization and genealogy of Resident Killer Whales (Orcinus orca) in the coastal waters of British Columbia and Washington State. Report of the International Whaling Commission Special Issue 12:383-405. Bigg, M.A., G.M. Ellis, J.K.B. Ford, and K.C. Balcomb. 1987. Killer whales: a study of their identification, genealogy, and natural history in British Columbia and Washington State. Phantom Press, Nanaimo, British Columbia. Bigg, M.A. and A.A. Wolman. 1975. Live-capture Killer Whales (Orcinus orca) fishery, British Columbia and Washington, 1962-73. Journal of the Fisheries Research Board of Canada 32: 1213-1221. Bigler, B.S., D.W. Welch, and J.H. Helle. 1996. A review of size trends among North Pacific salmon (Oncorhynchus spp.). Canadian Journal of Fisheries and Aquatic Science 53: 455-465. Black, N.R., A. Schulman-Janiger, R.L. Ternullo, and M. Guerrero-Ruiz. 1997. Killer Whales of California and western Mexico: a catalogue of photo-identified individuals. U.S. Dept. of Commerce NOAA-TM-NMFS-SWFSC-247. Borrell, A., D. Block, and G. Desportes. 1995. Age trends and reproductive transfer of organochlorine compounds in long-finned pilot whales from the Faroe Islands. Environmental Pollution 88: 283-292. Bottomley, J.A. and J. Theriault. 2003. DRDC Atlantic Q-273 Sea Trial Marine Mammal Impact Mitigation Plan. DRDC Atlantic TM 2003-044, Defence Research and Development Canada – Atlantic. Available at: http://cradpdf.drdc-rddc.gc.ca/PDFS/unc31/p522682.pdf [accessed February 2005].

258 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Branch, T.A. and D.S. Butterworth. 2001. Estimates of abundance south of 60oS for cetacean species sighted frequently on the 1978/79 to 1997/98 IWC/IDCR-SOWER sighting surveys. Journal of Cetacean Research and Management 3: 251-270. Briggs, D.A. 1991. Impact of human activities on Killer Whales at the rubbing beaches in the Robson Bight Ecological Reserve and adjacent waters during the summers of 1987 and 1989. Unpublished Report, BC Parks, Government of BC. Calambokidis, J., D.E. Bain and S.D. Osmek. 1998. Marine mammal research and mitigation in conjunction with air gun operation for the USGS "SHIPS" seismic surveys in 1998. Contract Report submitted to the Minerals Management Service. Available at: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.639.91 [accessed February 2005]. Cameron, W.M. 1941. Killer Whales stranded near Masset. Progress Report, Biological Station, Nanaimo, British Columbia and Pacific Experiment Station, Prince Rupert, British Columbia 49:16-17. Carl, G.C. 1946. A school of Killer Whales stranded at Estevan Point, Vancouver Island. Report of the Provincial Museum of Natural History and Anthropology B21-28. COSEWIC. 2008. COSEWIC assessment and update status report on the Killer Whale Orcinus orca, Southern Resident population, Northern Resident population, West Coast Transient population, Offshore population and Northwest Atlantic / Eastern Arctic population, in Canada. Committee on the Status of Endangered Wildlife in Canada. Ottawa. viii + 65 pp. Crum, L.A. and Y. Mao. 1996. Acoustically enhanced bubble growth at low frequencies and its implications for human diver and marine mammal safety. Journal of the Acoustical Society of America 99: 2898-2907. Cullon, D.L., M.B. Yunker, C. Alleyne, N.J. Dangerfield, S. O’Neill, M.J. Whiticar, and P.S. Ross. 2009. Persistent organic pollutants in Chinook salmon (Oncorhynchus tshawytscha): Implications for resident killer whales of British Columbia and adjacent waters. Environmental Toxicology and Chemistry 28:148–161. Dahlheim, M., D.K. Ellifrit, and J.D. Swenson. 1997. Killer Whales of southeast Alaska: a catalogue of photo-identified individuals. National Marine Mammal Laboratory, National Marine Fisheries Service, U.S. Department of Commerce, Seattle, Washington. Dahlheim, M.E. and J.E. Heyning. 1999. Killer Whale Orcinus orca (Linnaeus, 1758). Pages 281-322 in S. Ridgway and R. Harrison (editors). Handbook of marine mammals, Volume 6. Academic Press, San Diego, California. Dahlheim, M.E. and C.O. Matkin. 1994. Assessment of injuries to Prince William Sound Killer Whales. Pages 163-171 in T.R. Loughlin (editor). Marine mammals and the Exxon Valdez. Academic Press, San Diego, California. Darnerud, P.O. 2003. Toxic effects of brominated flame retardants in man and in wildlife. Environment International 29: 841-853. De Swart, R.L., P.S. Ross, J.G. Vos, and A.D.M.E. Osterhaus. 1996. Impaired immunity in Harbour Seals (Phoca vitulina) exposed to bioaccumulated environmental contaminants: review of a long-term study. Environmental Health Perspectives 104 (suppl. 4): 823-828. Deecke, V.B., J.K.B. Ford, and P. Spong. 2000. Dialect change in Resident Killer Whales: implications for vocal learning and cultural transmission. Animal Behaviour 60: 629-638.

259 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

DFO. 2005. Canada’s Policy on the Conservation of Wild Pacific Salmon. Vancouver, DFO Available at: http://www.pac.dfo-mpo.gc.ca/fm-gp/species-especes/salmon-saumon/wsp- pss/index-eng.html [accessed February 2018]. DFO. 2016a. Statement of Canadian Practice with respect to the Mitigation of Seismic Sound in the Marine Environment. Available at: http://www.dfo- mpo.gc.ca/oceans/publications/seismic-sismique/index-eng.html [accessed January 2018]. DFO. 2016b. Report on the Progress of Recovery Strategy Implementation for the Northern and Southern Resident Killer Whales (Orcinus orca) in Canada for the Period 2009 – 2014. Species at Risk Act Recovery Strategy Report Series. Fisheries and Oceans Canada, Ottawa. iii + 51 pp. DFO. 2017a. Action Plan for the Northern and Southern Resident Killer Whale (Orcinus orca) in Canada. Species at Risk Act Action Plan Series. Fisheries and Oceans Canada, Ottawa. v + 33 pp. DFO. 2017b. Identification of habitats of special importance to Resident Killer Whales (Orcinus orca) off the west coast of Canada. Canadian Science Advisory Secretariat Science Advisory Report 2017/11. DFO. 2017c. Southern Resident Killer Whale: A science based review of recovery actions for three at-risk whale populations. Available at: http://www.dfo-mpo.gc.ca/fm-gp/mammals- mammiferes/publications/whalereview-revuebaleine/review-revue/killerwhale- epaulard/index-eng.html [accessed February 2018]. DFO. 2018. Regulations Amending the Marine Mammal Regulations: SOR/2018-126. Available at:http://www.gazette.gc.ca/rp-pr/p2/2018/2018-07-11/html/sor-dors126-eng.html. [Accessed August 2018]. Donague, M., R. Reeves, and G.S. Andrews. 2002. Report of the Workshop On Interactions Between Cetaceans And Longline Fisheries. Apia, Samoa: November 2002. New England Aquarium Aquatic Forum Series Report 03-1, New England Aquarium, Boston, MA, USA. Duffield, D.A., D.K. Odell, J.F. McBain, and B. Andrews. 1995. Killer Whale (Orcinus orca) reproduction at Sea World. Zoo Biology 14:417-430. EIS. 2007. Final supplemental environmental impact statement for surveillance towed array sensor system low frequency active (SURTASS LFA) sonar. Available at: http:/www.surtass-lfa-eis.com/downloads/http:/www.surtass-lfa-eis.com/downloads/ [accessed March 2018]. Endo, T., Y. Hotta, K. Haraguchi, and M. Sakata. 2003. Mercury contamination in the red meat of whales and dolphins marketed for human consumption in Japan. Environmental Science and Technology 37: 2681-2685. Erbe, C. 2002. Underwater noise of whale-watching boats and potential effects on Killer Whales (Orcinus orca), based on an acoustic impact model. Marine Mammal Science 18:394- 418. Erbe, C. 2013. Underwater noise of small personal watercraft (jet skis). Journal of the Acoustical Society of America 133(4): EL326. European Parliament. 2004. European Session Document, Motion for a Resolution. B60089/04, Resolution P6 TA, 2004

260 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Evans, D.L. and G.R. England. 2001. Joint interim report Bahamas marine mammal stranding event of 15-16 March 2000. NOAA U.S. Dept. of Commerce and Dept. of the Navy.Available at: https://repository.library.noaa.gov/view/noaa/16198/noaa_16198_DS1 [accessed August]. 2018]. EVS (Environmental Consultants). 2003. Status, trends and effects of toxic contaminants in the Puget Sound environment: recommendations. Prepared for Puget Sound Action Team, Olympia WA. Available at: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.663.171 [accessed February 2005]. Ewald, G., P. Larsson, H. Linge, L. Okla, and N. Szarzi. 1998. Biotransport of organic pollutants to an inland Alaska lake by migrating sockeye salmon (Oncorhyncus nerka). Arctic 51: 40-47. Fay, R.R. and A.N. Popper. 2000. Evolution of hearing in vertebrates: the inner ears and processing. Hearing Research 149: 1-10. Felleman, F.L., J.R. Heimlich-Boran, and R.W. Osborne. 1991. The feeding ecology of Killer Whales (Orcinus orca) in the Pacific Northwest. Pages 113-147 in K. Pryor and K.S. Norris (editors). Dolphin societies: discoveries and puzzles. University of California Press, Berkeley, California. Ferguson, S.H., J.W. Higdon, and E.G. Chmelnitsky. 2010. The rise of Killer Whales as a major Arctic predator. In A little less Arctic (pp. 117-136). Springer, Dordrecht. Fernandez, A., M. Arbelo, R. Deaville, I.A.P. Patterson, P. Castro, J.R. Baker, R. Degolloada, H.M. Ross, P, Herraez, A.M. Pocknell, E. Rodrigez, F.E. Howie, A. Espinosa, R.J. Reid, R. Jaber, V. Martin, A.A. Cunningham, and P.D. Jepson. 2004. Whales, sonar and decompression sickness (reply). Nature 428: 1-2. Finneran, J.J. 2003. Whole lung resonance in a bottlenose dolphin (Tursiops truncatus) and white whale (Delphinapterus leucas). Journal of the Acoustical Society of America 114: 529-535. Fleeger, J.W., K.R. Carman and R.M. Nisbet. 2003. Indirect effects of contaminants in aquatic ecosystems. The Science of the Total Environment 317: 207-233. Foote, A.D., R.W. Osborne, and A. R. Hoelzel. 2004. Whale-call response to masking boat noise. Nature 428: 910. Ford, J.K.B. 1984. Call traditions and dialects of Killer Whales (Orcinus orca) in British Columbia. Ph.D. dissertation, University of British Columbia. 435 p. Ford, J.K.B. 1989. Acoustic behaviour of Resident Killer Whales Orcinus orca off Vancouver Island, British Columbia, Canada. Canadian Journal of Zoology 67: 727-745. Ford, J.K.B. 1991. Vocal traditions among Resident Killer Whales, Orcinus orca, in coastal waters of British Columbia, Canada. Canadian Journal of Zoology 69: 1454-1483. Ford, J.K.B. 2006. An Assessment of critical habitats of Resident Killer Whales in waters off the Pacific Coast of Canada. Canadian Science Advisory Secretariat, Research Document 2006/072. Ford, J.K.B. 2014. Marine Mammals of British Columbia. Royal BC Museum Handbook, Mammals of BC, volume 6. Royal BC Museum, Victoria. 460 pp.

261 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Ford, J.K.B., pers. comm. 2018. Former Head of Marine Mammal Research, Fisheries and Oceans Canada, Pacific Biological Station, Nanaimo, British Columbia. Ford, J.K.B. and G.M. Ellis. 1999. Transients: Mammal-hunting Killer Whales. UBC Press, Vancouver, British Columbia. 96 pp. Ford, J.K.B. and G.M. Ellis. 2002. Reassessing the social organization of Resident Killer Whales in British Columbia. Pages 72-75 in the Fourth International Orca Symposium and Workshop, September 23-28, 2002. CEBC-CNRS, France. Ford, J.K.B. and G.M. Ellis. 2005. Prey selection and food sharing by fish-eating ‘Resident’ Killer Whales (Orcinus orca) in British Columbia. Canadian Science Advisory Secretariat, Research Document 2005/041. Ford, J.K.B. and G.M. Ellis. 2006. Selective foraging by fish-eating killer whales Orcinus orca in British Columbia. Marine Ecology Progress Series 316: 185-199. Ford, J.K.B., G.M. Ellis, and K.C. Balcomb. 2000. Killer Whales: the natural history and genealogy of Orcinus orca in British Columbia and Washington, second edition. UBC Press, Vancouver, British Columbia. 104 pp. Ford, J.K.B., G.M. Ellis, L.G. Barrett-Lennard, A.B. Morton, R.S. Palm, and K.C. Balcomb. 1998. Dietary specialization in two sympatric populations of Killer Whales (Orcinus orca) in coastal British Columbia and adjacent waters. Canadian Journal of Zoology 76:1456- 1471. Ford, J.K.B., G.M. Ellis, and P.F. Olesiuk. 2005. Linking prey and populations dynamics: did food limitation cause recent declines of ‘Resident’ Killer Whales (Orcinus orca) in British Columbia. Canadian Science Advisory Secretariat, Research Document 2005/042. Ford, J.K.B., G.M. Ellis, P.F. Olesiuk, and K.C. Balcomb. 2010. Linking Killer Whale survival and prey abundance: food limitation in the oceans’ apex predator? Biology letters 6:139–42. Ford, J.K.B., J.F. Pilkington, A. Riera, M. Otsuki, B. Gisborne, R.M. Abernethy, E.H. Stredulinsky, J.R. Towers, and G.M. Ellis. 2017. Habitats of special importance to Resident Killer Whales (Orcinus orca) off the west coast of Canada. DFO Can. Sci. Advis. Sec. Res. Doc. 2017/035. viii + 57 p. Ford, J.K.B, E.H. Stredulinsky, G.M. Ellis, J.W. Durban, and J.F. Pilkington. 2014. Offshore Killer Whales in Canadian Pacific Waters: Distribution, Seasonality, Foraging Ecology, Population Status and Potential for Recovery. DFO Canadian Science Advisory Secretariat Research Document 2014/088. vii + 55 p. Ford, J.K.B, E.H. Stredulinsky, J.R. Towers, and G.M. Ellis. 2013. Information in support of the identification of critical habitat for Transient Killer Whales (Orcinus orca) off the west coast of Canada. DFO Canadian Science Advisory Secretariat Research Document 2012/155. iv + 46 p. Ford, M.J., K.M. Parsons, E.J. Ward, J.A. Hempelmann, C.K. Emmons, M.B. Hanson, K.C. Balcomb, and L.K. Park. 2018. Inbreeding in an endangered Killer Whale population. Animal Conservation, doi: 10.1111/acv 12413. Forney, K. A. and P. Wade. 2006. Worldwide distribution and abundance of Killer Whales. In J.A. Estes, R.L. Brownell, Jr., D. P. DeMaster, D.F. Doak, and T.M. Williams (editors). Whales, whaling and ocean ecosystems. University of California Press, Berkeley, California.

262 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Foster, E.A., D.W. Franks, S. Mazzi, S.K. Darden, K.C. Balcomb, J.K.B. Ford, and D.P. Croft. 2012. Adaptive prolonged postreproductive life span in killer whales. Science 337:1313. Francis, R.C., S.R. Hare, A.B. Hallowed and W.S. Wooster. 1998. Effects of interdecadal climate variability on the oceanic ecosystems of the NE Pacific. Fisheries Oceanography 7: 1-21. Freeman, D., pers. comm. 2007. Department of National Defence. Gardner, J., D.L. Peterson, A. Wood, and V. Maloney. 2004. Making sense of the debate about hatchery impacts: interactions between enhanced and wild salmon on Canada’s Pacific coast. Prepared for the Pacific Fisheries Resource Conservation Council, Vancouver, BC. 187 pp. Available at: https://www.psf.ca/sites/default/files/lib_195.pdf [accessed December 18, 2004] Garrett and Ross. 2010. Recovering Resident Killer Whales: A guide to contaminant sources, mitigation, and regulations in British Columbia. Canadian Technical Report of Fisheries and Aquatic Sciences 2894: xiii + 224. Gaydos, J.K., K.C. Balcomb, R.W. Osborne, and L. Dierauf. 2004. Evaluating potential infectious disease threats for Southern Resident Killer Whales, Orcinus orca: a model for endangered species. Biological Conservation 117: 253-262. Geraci, J.R., D.M. Anderson, R.J. Timperi, D.J. St. Aubin, G.A. Early, J.H. Prescott, and C.A. Mayo. 1989. Humpback whales (Megaptera novaeangliae) fatally poisoned by dinoflagellate toxin. Canadian Journal of Fisheries and Aquatic Science 46: 1895-1898. Geraci, J.R. and D.J. St. Aubin. 1982. Sea mammals and oil: confronting the risks. Academic Press, New York, New York. Grachev, M.A., V.P. Kumarev, and L.V. Mamaev. 1989. Distemper virus in Baikal seals. Nature 338: 209. Grant, S.C.H. and P.S. Ross. 2002. Southern Resident Killer Whales at risk: toxic chemicals in the British Columbia and Washington environment. Canadian Technical Report of Fisheries and Aquatic Science 2412:1-111. Greenwood, A.G. and D.C. Taylor. 1985. Captive Killer Whales in Europe. Aquatic Mammals 1:10-12. Guenther, T.J., R.W. Baird, R.L Bates, P.M. Willis, R.L. Hahn, and S.G. Wischniowski. 1995. Strandings and fishing gear entanglements of cetaceans on the west coast of Canada in 1994. Report submitted to the International Whaling Commission, SC/47/06. Haenel, N.J. 1986. General notes on the behavioural ontogeny of Puget Sound Killer Whales and the occurrence of allomaternal behaviour. Pages 285-300 in B.C. Kirkevold and J.S. Lockard (editors). Behavioural Biology of Killer Whales. Alan R. Liss, New York, New York. Hagen, M.E., A.G. Colodey, W.D. Knapp, and S.C. Samis. 1997. Environmental response to decreased dioxin and furan loadings from British Columbia coastal pulp mills. Chemosphere 34: 1221-1229. Haggarty, D.R., B. McCorquodale, D.I. Johannessen, C.D. Levings, and P.S. Ross. 2003. Marine environmental quality in the central coast of British Columbia, Canada: A review of contaminant sources, types and risks. Canadian Technical Report of Fisheries and Aquatic Sciences 2507. 153 pp.

263 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Hall, J.D. and C.S. Johnson. 1972. Auditory thresholds of a Killer Whale Orcinus orca Linnaeus. Journal of the Acoustical Society of America 51: 515-517. Hall, A.J., O.I. Kalantzi, and G.O. Thomas. 2003. Polybrominated diphenyl ethers (PBDEs) in Grey Seals during the first year of life – are they thyroid hormone endocrine disrupters? Environmental Pollution 126: 29-37. Hanson, M.B., R.W. Baird, J.K.B. Ford, J. Hempelmann-Halos, D.M. Van Doornik, J.R. Candy, C.K. Emmons, G.S. Schorr, B. Gisborne, K.L. Ayres, S.K. Wasser, K.C. Balcomb, J.G. Sneva, and M.J. Ford. 2010. Species and stock identification of prey consumed by endangered Southern Resident Killer Whales in their summer range. Endangered Species Research 11: 69–82. Hanson, M.B., C.K. Emmons, E.J. Ward, J.A. Nystuen, and M.O. Lammers. 2013. Assessing the coastal occurrence of endangered killer whales using autonomous passive acoustic recorders. Journal of the Acoustical Society of America 134:3486-3495. Hartwell, S.I. 2004. Distribution of DDT in sediments off the central California coast. Marine Pollution Bulletin 49: 299-305. Healey, M.C. 1991. Life history of Chinook Salmon (Oncorhynchus tshawytscha). Pages 313- 393 in C. Groot and L. Margolis (editors). Pacific Salmon life histories. UBC Press, Vancouver, British Columbia. Heimlich-Boran, J.R. 1986. Fishery correlations with the occurrence of Killer Whales in Greater Puget Sound. Pages 113-131 in B.C. Kirkevold and J.S. Lockard (editors). Behavioural biology of Killer Whales. Alan R. Liss, New York, New York. Heise, K., L. G. Barrett-Lennard, E. Saulitis, C. Matkin, and D. Bain. 2003. Examining the evidence for Killer Whale predation on Steller sea lions in British Columbia and Alaska. Aquatic Mammals 29: 325-334. Heyning, J.E. and M.E. Dahlheim. 1988. Orcinus orca. Mammalian Species 304: 1-9. Heyning, J.E., T.D. Lewis, and C.D. Woodhouse. 1994. A note on odontocete mortality from fishing gear entanglements off southern California. Reports of the International Whaling Commission Special Issue 15: 439-442. Hoelzel, A.R., A. Natoli, M.E. Dahlheim, C. Olavarria, R.W. Baird, and N.A. Black. 2002. Low worldwide genetic diversity in the Killer Whale (Orcinus orca): implications for demographic history. Proceedings of the Royal Society of London, Biological Sciences, Series B. 269:1467-1473. Holt, M.M., D.P. Noren, V. Veirs, C.K. Emmons, and S. Veirs. 2009. Speaking up: killer whales (Orcinus orca) increase their call amplitude in response to vessel noise. Journal of the Acoustical Society of America 125(1): 27-32. Hooper, K. and T.A. McDonald. 2000. The PBDEs: an emerging environmental challenges and another reason for breast-milk monitoring programs. Environmental Health Perspectives 108: 387- 392. IUCN. 2004. RESWCC3.068 Undersea noise pollution Congress reference: GR3.RES053.Rev1. Available at: https://awionline.org/sites/default/files/uploads/documents/IUCN_RES053-1238105850- 10131.pdf [accessed February 2005].

264 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

IWC (International Whaling Commission). 2004. Annex K. Report of the Standing Working Group on Environmental Concerns. Report of the Scientific Committee of the International Whaling Commission. Meeting held in Sorrento Italy, 29 June – 10 July 2004. Jelinski, D.E., C.C. Krueger, and D.A. Duffus. 2002. Geostatistical analyses of interactions between Killer Whales (Orcinus orca) and recreational whale-watching boats. Applied Geography 22: 393-411. Jepson, P.D., C.R. Allchin, R.J. Law, T. Kuiken, J.R. Baker, E. Rogan, and J.T. Kirkwood. 1999. Investigating potential associations between chronic exposure to polychlorinated biphenyls and infectious disease mortality in harbour porpoises from England and Wales. Science of the Total Environment 243-244: 339-348. Jepson, P.D., M. Arbelo, R. Deaville, I.A.P. Patterson, P. Castro, J.R. Baker, E. Degollada, H.M. Ross, P. Herráez, A.M. Pocknell, F. Rodríguez, F.E. Howie, A. Espinosa, R.J. Reid, J.R. Jaber, V. Martin, A.A. Cunningham, and A. Fernández. 2003. Gas bubble lesions in stranded cetaceans. Nature 425: 575. Kannan, K., J. Koistinen, K. Beckmen, T. Evans, J.F. Gorzelany, K.J. Hansen, P.D. Jones, E. Helle, M. Nyman, and J.P. Giesy. 2001. Accumulation of perfluorooctane sulfonate in marine mammals. Environmental Science and Technology 35: 1593-1598. Kennedy, S., T. Kuiken, and P.D. Jepson. 2000. Mass die-off of Caspian seals caused by canine distemper virus. Emerging Infectious Diseases 6: 637-639. Ketten, D.R., J. Lien, and S. Todd. 1993. Blast injury in humpback whales: evidence and implications. Journal of the Acoustical Society of America 94: 1849-1850. Krahn, M.M., M.J. Ford, W.F. Perrin, P.R. Wade, R.P. Angliss, M.B. Hanson, B.L. Taylor, G.M. Ylitalo, M.E. Dahlheim, J.E. Stein, and R.S. Waples. 2004. 2004 Status review of Southern Resident Killer Whales (Orcinus orca) under the Endangered Species Act. U.S. Dept. Commer, NOAA Tech. Memo. NMFSNWFSC- 62, 73 p. Available at: http://www.westcoast.fisheries.noaa.gov/protected_species/marine_mammals/killer_wha le/ [accessed January 2005]. Krahn, M.M., M.B. Hanson, G.S. Schorr, C.K. Emmons, D.G. Burrows, J.L Bolton, R.W. Baird, and G.M. Ylitalo. 2009. Effects of age, sex and reproductive status on persistent organic pollutant concentrations in “Southern Resident” Killer Whales. Marine Pollution Bulletin 58:1522–1529. Krahn, M.M., P.R. Wade, S.T. Kalinowski, M.E. Dahlheim, B.L. Taylor, M.B. Hanson, G.M. Ylitalo, R.P. Angliss, J.E. Stein, and R.S. Waples. 2002. Status review of Southern Resident Killer Whales (Orcinus orca) under the Endangered Species Act. U.S. Department of Commerce, NOAA Technical Memorandum NMFS-NWFSC-54. 133p. Kriete, B. 1995. Bioenergetics in the Killer Whale, Orcinus orca. Ph.D. Thesis, University of British Columbia, Vancouver, British Columbia. Kriete, B. 2002. Bioenergetic changes from 1986 to 2002 in Southern Resident Killer Whales (Orcinus orca). Page 88 in Fourth international orca symposium and workshops, September 23-28, 2002, CEBC-CNRS, France. Kruse, S. 1991. The interactions between Killer Whales and boats in Johnstone Strait, B.C. Pages 149-159 in K. Pryor and K.S. Norris, editors. Dolphin societies: discoveries and puzzles. University of California Press, Berkeley, California.

265 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Lacy, R.C., R. Williams, E. Ashe, K.C. Balcomb III, L.J.N. Brent, C.W. Clark, D.P. Croft, D.A. Giles, M. MacDuffee, and P.C. Paquet. 2017. Evaluating anthropogenic threats to endangered Killer Whales to inform recovery plans. Scientific Reports 7:14119. Langelier, K.M., P.J. Stacey, and R.W. Baird. 1990. Stranded whale and dolphin program of BC – 1989 report. Wildlife Veterinary Report 3(1): 10-11. Law, R.J., A. Covaci, S. Herrad, D. Herzke, M.A.-E. Abdallah, K. Fernie, L.-M.L. Toms, H. Takigami. 2014. Levels and trends of PBDEs and HBCDs in the global environment: status at the end of 2012. Environment International 65:147-158. Levings, C.D., J.R. Cordell, S. Ong, and G.E. Piercey. 2004. The origin and identify of invertebrate organisms being transported to Canada’s Pacific coast by ballast water. Canadian Journal of Fisheries and Aquatic Sciences 61: 1-11. Lindstrom, G., H. Wingfors, M. Dam, and B. von Bavel. 1999. Identification of 19 polybrominated diphenyl ethers (PBDEs) in long-finned pilot whale (Globicephala melas) from the Atlantic. Archives of Environmental Contamination & Toxicology 36: 355-363. Ljungblad, D.K.B., B. Wursig, S.L. Swartz and J.M. Keene. 1988. Observations on the behavioural responses of bowhead whales (Balaena mysticetus) to active geophysical vessels in the Alaskan Beaufort Sea. Arctic 41: 183-194. Lusseau, D., D. Bain, R. Williams, and J. Smith. 2009. Vessel traffic disrupts the foraging behavior of Southern Resident Killer Whales Orcinus orca. Endangered Species Research 6:211–221.Madsen, P.T., B. Mohl, B.K. Nielsen, and M. Wahlberg. 2002. Male sperm whale behaviour during exposures to distant seismic survey pulses. Aquatic Mammals 28: 231-240. Mahnken, C., G. Ruggerone, W. Wakinitz, and T. Flagg. 1998. A historical perspective on salmonid production from Pacific Rim hatcheries. North Pacific Anadromous Fish Commission Bulletin 1: 38-53. Malme, C.I. and P.R. Miles 1987. The influence of sound propagation conditions on the behavioural responses of whales to underwater industrial noise. Journal of the Acoustical Society of America 1: 97. Malme, C.I., P.R. Miles, P. Tyack, C.W. Clark, and J.E. Bird. 1985. Investigation of the potential effects of underwater noise from petroleum industry activities on feeding humpback whale behaviour. BBN Rep. 5851: OCS Study MMS 85-0019 for UUUUU.S. Minerals Management Service, Anchorage, Alaska. Martoja, R. and R.J. Berry. 1980. Identification of tiemannite as a probable product of demethylation of mercury by selenium in cetaceans, a complement scheme of the biological cycle of mercury. Vie Milieu 30: 7-10. Matkin, C.O., G.M. Ellis, E.L. Saulitis, L.G. Barrett-Lennard, and D.R. Matkin. 1999. Killer Whales of Southern Alaska. North Gulf Oceanic Society, Homer, Alaska. Matkin, C.O. and E.L. Saulitis. 1994. Killer Whale (Orcinus orca) biology and management in Alaska. Report for the Marine Mammal Commission, Washington, D.C.

Matkin, C.O., E.L. Saulitis, G.M. Ellis, P. Olesiuk, and S.D. Rice. 2008. Ongoing population-level impacts on Killer Whales Orcinus orca following the ‘Exxon Valdez’ oil spill in Prince William Sound, Alaska. Mar. Ecol. Prog. Ser. 356:269-281.

266 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

McCallum, H., D. Harvell, and A. Dobson. 2003. Rates of spread of marine pathogens. Ecological Letters 6: 1062-1067. McCauley, R.D., J. Fewtrell, and A.N. Popper. 2003. High intensity anthropogenic sound damages fish ears. Journal of the Acoustic Society of America 113:638-642. Mikhalev, Y.A., M.V. Ivashin, V.P. Savusin, and F.E. Zelenaya. 1981. The distribution and biology of Killer Whales in the Southern hemisphere. Report of the International Whaling Commission 31:551-566 Miller, P.O. and D.E. Bain. 2000. Within-pod variation in the sound production of a pod of Killer Whales, Orcinus orca. Animal Behaviour 60: 617-628. Mitchell, E. and R.R. Reeves. 1988. Records of Killer Whales in the western North Atlantic, with emphasis on eastern Canadian waters. Rit Fiskideildar 11:161-193. Morton, A.B. and H.K. Symonds. 2002. Displacement of Orcinus orca (L.) by high amplitude sound in British Columbia, Canada. ICES Journal of Marine Science 59:71-80. Mos, L., P.S. Ross, D. McIntosh, and S. Raverty. 2003. Canine distemper virus in River Otters in British Columbia as an emergent risk for coastal pinnipeds. Veterinary Record 152: 237-239. Muir, D., B. Braune, B. De March, R. Norstrom, R. Wagemann, L. Lockhart, B. Hargrave, D. Bright, R. Addison, J. Payne, and K. Reimer. 1999. Spatial and temporal trends and effects of contaminants in the Canadian Arctic marine ecosystem: a review. Science of the Total Environment 230: 83-144. Myrberg Jr., A.A. 1990. The effects of man-made noise on the behaviour of marine animals. Environment International 16: 575-586. Nichol, L.M. and D.M. Shackleton. 1996. Seasonal movements and foraging behaviour of Northern Resident Killer Whales (Orcinus orca) in relation to the inshore distribution of salmon (Oncorhynchus spp.) in British Columbia. Canadian Journal of Zoology 74: 983- 991. Nieukirk, S.L., K.M. Stafford, D.K. Mellinger, R.P. Dziak, and C.G. Fox. 2004. Low-frequency whale and seismic airgun sounds recorded in the mid-Atlantic Ocean. Journal of the Acoustical Society of America 115: 1832-1843. NIMMSA. 2016. NIMMSA Code of Conduct. Available at: http://nimmsa.org/code-of-conduct/ [accessed January 2018]. Nishiwaki, M. and C. Handa. 1958. Killer Whales caught in the coastal waters off Japan for recent 10 years. Scientific Report of the Whales Research Institute, Tokyo 13: 85-96. NMFS (National Marine Fisheries Service). 2004. Preliminary report: multidisciplinary investigation of Harbour Porpoises (Phocoena phocoena) stranded in Washington State from 2 May-2 June 2003 coinciding with the mid-range sonar exercises of the USS Shoup, February, 2004. 109 pp. NMFS. 2005. Endangered Fish and Wildlife; Notice of Intent to Prepare an Environmental Impact Statement. Federal Register (Vol. 70, Num. 7) January 11, 2005. pp 1871-1875. NMFS. 2006a. Designation of critical habitat for Southern Resident Killer Whales, biological report, October 2006, 44 pp. Available at:

267 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

http://www.westcoast.fisheries.noaa.gov/publications/protected_species/marine_mamma ls/killer_whales/esa_status/srkw-ch-bio-rpt.pdf [accessed May 2018]. NMFS. 2006b. Endangered and Threatened Species: Designation of Critical Habitat for Southern Resident Killer Whale, Federal Register, 71 FR 69054. NMFS. 2016. Technical guidance for assessing the effects of anthropogenic sound on marine mammal hearing: Underwater acoustic thresholds for onset of permanent and temporary threshold shifts. U.S. Department of Commerce, NOAA. NOAA Technical Memorandum NMFS-OPR-55, 178 p. NOAA. 2016. Southern Resident Killer Whale tagging. Available at: https://www.nwfsc.noaa.gov/research/divisions/cb/ecosystem/marinemammal/satellite_t agging/index.cfm [Accessed February 2018]. NOAA. 2017. West Coast Salmon & Steelhead Listings. Available at: http://www.westcoast.fisheries.noaa.gov/protected_species/salmon_steelhead/salmon_a nd_steelhead_listings/salmon_and_steelhead_listings.html [Accessed January 2018]. Northcote, T.G. and P.A. Larkin. 1989. The Fraser River: A major salmonine production system. P.172-204 IN: Dodge, D.P. (ed.). Proceedings of the International Large River Symposium. Can. Spec. Publ. in Fish. Aquat. Sci. 106. Nowacek, D.P., L.H. Thorne, D.W. Johnston, and P.L. Tyack. 2007. Responses of cetaceans to anthropogenic noise. Mammal Review 32:81-115. NRC (National Research Council). 2003. Ocean Noise and Marine Mammals. National Research Council, National Academies Press, Washington, D.C. NRC. 2005. Marine Mammal Populations and Ocean Noise: Determining When Noise Causes Biologically Significant Effects. National Academies Press, Washington, D.C. Nyman, M., M. Bergknut, M.L. Fant, H. Raunio, M. Jestoi, C. Bengs, A. Murk, J. Koistinen, C. Bäckman, O. Pelkonen, M. Tysklind, T. Hirvi, and E. Helle. 2003. Contaminant exposure and effects in Baltic ringed and grey seals as assessed by biomarkers. Marine Environmental Research 55: 73-99. Øien, N. 1988. The distribution of Killer Whales (Orcinus orca) in the North Atlantic based on Norwegian catches, 1938-1981, and incidental sightings, 1967-1987. Rit Fiskideildar 11: 65-78. Olesiuk, P.F., M.A. Bigg, and G.M. Ellis. 1990. Life history and population dynamics of Resident Killer Whales (Orcinus orca) in the coastal waters of British Columbia and Washington State. Report of the International Whaling Commission Special Issue 12: 209-243. Olesiuk, P.F., L.M. Nichol, M.J. Sowden, and J.K.B. Ford. 2002. Effect of the sound generated by an acoustic harassment device on the relative abundance and distribution of Harbour Porpoises (Phocoena phocoena) in Retreat Passage, British Columbia. Marine Mammal Science 18: 843-862. O’Neill, S.M., J.E. West, and J.C. Hoeman. 1998. Spatial trends in the concentration of polychlorinated biphenyls (PCBs) in Chinook (Oncorhynchus tshawytscha) and Coho Salmon (O. kisutch) in Puget Sound and factors affecting PCB accumulation: results from the Puget Sound Ambient Monitoring Program. Puget Sound Research ’98:312- 328. Available at: https://wdfw.wa.gov/publications/01031/ [accessed March 2018].

268 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Osborne, R.W. 1991. Trends in Killer Whale movements, vessel traffic, and whale watching in Haro Strait. Pages 672-688 in Puget Sound Research ’91 Proceedings. Puget Sound Water Quality Authority, Olympia, Washington. Osborne, R.W. 1999. A historical ecology of Salish Sea “Resident” Killer Whales (Orcinus orca): with implications for management. Ph.D. Thesis, University of Victoria, Victoria, British Columbia. Osborne, R.W., K. Koski, and R. Otis. 2003. Trends in whale watching traffic around Southern Resident Killer Whales. PowerPoint presentation from The Whale Museum, Friday Harbor, Washington. Parks Canada Agency. 2016. Multi-species Action Plan for Gwaii Haanas National Park Reserve, National Marine Conservation Area Reserve, and Haida Heritage Site. Species at Risk Act Action Plan Series. Parks Canada Agency, Ottawa. vi+ 25 pp. Parks Canada Agency. 2017. Multi-species Action Plan for Pacific Rim National Park Reserve of Canada - Species at Risk Act Action Plan Series. Parks Canada Agency, Ottawa. v + 29 pp. Parks Canada Agency. 2018. Multi-species Action Plan for Gulf Islands National Park Reserve of Canada. Species at Risk Act Action Plan Series. Parks Canada Agency, Ottawa. vi+ 27 pp. Perrin, W.F. and J.R. Geraci. 2002. Stranding. Pages 1192-1197 in W.F. Perrin, B. Würsig, and J.G.M. Thewissen (editors). Encyclopedia of marine mammals. Academic Press, San Diego, California. Piantadosi, C.A. and E.D. Thalman. 2004. Whales, sonar and decompression sickness. Nature 428: 1. Pike, G.C. and I.B. MacAskie. 1969. Marine mammals of British Columbia. Bulletin of the Fisheries Research Board of Canada 171: 1-54. Pitman, R.L. and P. Ensor. 2003. Three forms of Killer Whales (Orcinus orca) in Antarctic waters. Journal of Cetacean Research and Management 5: 131-139. PWWA. 2017. Pacific Whale Watch Association Whale Watching Guidelines. Available at: https://www.pacificwhalewatchassociation.com/guidelines/ [accessed January 2018]. Raverty, S.A., J.K. Gaydos, and J.A. St. Leger. 2014. Killer Whale necropsy and disease testing protocol. Available at: https://static1.squarespace.com/static/5b071ddea2772cebc1662831/t/5b29e3bb352f539 b40739ecd/1529471951000/Orca-necropsy-protocol-FINAL-May-15-2014.pdf [accessed March 2018]. Rayne, S., M.G. Ikonomou, P.S. Ross, G.M. Ellis, and L.G. Barrett-Lennard. 2004 PBDEs, PBBs, and PCNs in three communities of free-ranging Killer Whales (Orcinus orca) from the northeastern Pacific Ocean. Environmental Science and Technology 38: 4293-4299. Reeves, R.R. and E. Mitchell. 1988. Distribution and seasonality of Killer Whales in the eastern Canadian Arctic. Fis Riskideildar 11:136-160. Reeves, R.R., W.F. Perrin, B.L. Taylor, C.S. Baker, and S.L. Mesnick. 2004. Report of the Workshop on Shortcomings of Cetacean Taxonomy in Relation to Needs of Conservation and Management April 30-May 2, 2004, La Jolla, California. NOAA

269 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Technical Memorandum NOAA-NMFS-SWFSC-363. 94 pp. Available at: https://repository.library.noaa.gov/view/noaa/3474 [accessed January 2005]. Reijnders, P.J.H. 1986. Reproductive failure in common seals feeding on fish from polluted coastal waters. Nature 324: 456-457. Rendell, L. and H. Whitehead. 2001. Culture in whales and dolphins. Behavioural and Brain Sciences 24: 309-382. Richardson, W.J., C.R. Greene, Jr., C.I. Malme, and D.H. Thomson. 1995. Marine mammals and noise. Academic Press, San Diego, California. Riddell, B.E. 1993. Spatial organization of Pacific salmon: what to conserve? Pages 22-41 in Genetic Conservation of Salmonid Fishes. J.G. Cloud and G.H. Thorgaard (editors). Plenum Press, New York. Riddell, B. 2004. Pacific salmon resources in central and north coast British Columbia. Prepared for the Pacific Fisheries Resource Conservation Council, Vancouver, BC. 23 pp. Available at: https://www.watershed- watch.org/publications/files/SalmonResources_2004.pdf [accessed December 2004]. Riera, A. 2012. Patterns of seasonal occurrence of sympatric Killer Whale lineages in transboundary waters off Southern Vancouver Island and Washington state, as determined by passive acoustic monitoring. MSc thesis, University of Victoria, Victoria, British Columbia. Ross, P.S. 2000. Marine mammals as sentinels in ecological risk assessment. Humans and Ecological Risk Assessment 6: 29-46. Ross, P.S. 2002. The role of immunotoxic environmental contaminants in facilitating the emergence of infectious diseases in marine mammals. Humans and Ecological Risk Assessment 8: 277-292. Ross, P.S. 2006. Fireproof Killer Whales (Orcinus orca): Flame retardant chemicals and the conservation imperative in the charismatic icon of British Columbia, Canada. Canadian Journal of Fisheries and Aquatic Sciences 63: 224-234. Ross, P.S., C.M. Couillard, M.G. Ikonomou, S.C. Johannessen, M. Lebeuf, R.W. Macdonald, and G.T. Tomy. 2009. Large and growing environmental reservoirs of Deca-BDE present an emerging health risk for fish and marine mammals. Marine Pollution Bulletin 58: 7–10. Ross, P.S., G.M. Ellis, J.K.B. Ford, and L.G. Barrett-Lennard. 2002. Toxic chemical pollution and Pacific Killer Whales (Orcinus orca). Pages 126-130 in Fourth International Orca Symposium and Workshops, September 23-28, 2002, CEBC-CNRS, France. Ross, P.S., G.M. Ellis, M.G. Ikonumou, L.G. Barrett-Lennard and R.F. Addison. 2000. High PCB concentrations in free-ranging Pacific Killer Whales, Orcinus orca: effects of age, sex and dietary preference. Marine Pollution Bulletin 40: 504-515. Ross, P.S., S.J. Jeffries, M.B. Yunker, R.E. Addison, M.G. Ikonomou, and J.C. Calambokidas. 2004. Harbour Seals (Phoca vitulina) in British Columbia, Canada, and Washington State, USA, reveal a combination of local and global polychlorinated byphenyl, dioxin and furan signals. Environmental Toxicology and Chemistry 23: 157-165. Ross, P.S., M. Noël, D. Lambourn, N. Dangerfield, J. Calambokidis, and S. Jeffries. 2013. Declining concentrations of persistent PCBs, PBDEs, PCDEs, and PCNs in Harbour Seals (Phoca vitulina) from the Salish Sea. Progress in Oceanography 115:160–170.

270 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Saulitis, E., C.O. Matkin and G. Ellis. 2002. The biology and status of an endangered Transient Killer Whale population in Prince William Sound, Alaska. Pages 131-132 in Fourth International Orca Symposium and Workshops, September 23-28, 2002, CEBC-CNRS, France. Scammon, C.M. 1874. The marine mammals of the northwestern coast of North America, together with an account of the American whale fishery. J.H. Carmany and Company, San Francisco, California. Scholin, C.A., F. Gulland, G.J. Doucette, S. Benson, M. Busman, F.P. Chavez, J. Cordaro, R. DeLong, A.D. Vogelaere, J. Harvey, M. Haulena, K. Lefebvre, T. Lipscomb, S. Loscutoff, L.J. Lowenstine, R. Marin, P.E. Miller, W.A. McLellan, P.D.R. Moeller, C.L. Powell, T. Rowles, P. Silvagni., M. Silver, T. Spraker, V. Trainer, and F.M.V. Dolah. 2000. Mortality of sea lions along the central California coast linked to a toxic diatom bloom. Nature 403:80-84. Seely, E., R.W. Osborne, K. Koski, and S. Larson. 2017. Soundwatch: Eighteen years of monitoring whale watch vessel activities in the Salish Sea. PLoS ONE 12(12): e0189764. Shore, V. 1995. Do Killer Whales get trapped often? Blackfish Sounder (Vancouver Aquarium) 3:3. Shore, V. 1998. Southern Residents go a ‘bridge too far’. Blackfish Sounder (Vancouver Aquarium) 6:3. Sih, A., A.M. Bell, and J.L. Kerby. 2004. Two stressors are far deadlier than one. Trends in Ecology and Evolution 19: 274-276. Smith, J.C. and D.E. Bain. 2002. Theodolite study of the effects of vessel traffic on Killer Whales (Orcinus orca) in the near-shore waters of Washington State, USA. Pages 143-145 in Fourth international orca symposium and workshops, September 23-28, 2002, CEBC- CNRS, France. Song, L., A. Seeger, and J. Santos-Such. 2005. On membrane motor activity and chloride flux in the outer hair cell: lessons learned from the environmental toxin tributyltin. Biophysical Journal 88 (3): 2350-2362. Stone C.J. 2003. The effects of seismic activity on marine mammals in UK waters, 1998-2000. Joint Nature Conservation Committee Report No. 323. Aberdeen, UK. Stredulinsky, E.H. 2016. Determinants of group splitting: an examination of environmental, demographic, genealogical, and state-dependent factors of matrilineal fission in a threatened population of fish-eating killer whales (Orcinus orca). MSc thesis, University of Victoria, Victoria, British Columbia. Syzmanski, M.D., D.E. Bain, K. Kiehl, S. Pennington, S. Wong and K.R. Henry. 1999. Killer Whale (Orcinus orca) hearing: Auditory brainstem response and behavioral audiograms. Journal of the Acoustical Society of America 106: 1134-1141. Tanabe, S. and R. Tatsukawa. 1992. Chemical modernization and vulnerability of cetaceans: increasing toxic threat of organochlorine contaminants. Pages 161-177 in C.H. Walker and D.R. Livingstone (editors). Persistent Pollutants in Marine Ecosystems. Pergamom Press, New York, New York.

271 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Taylor, M. and B. Plater. 2001. Population viability analysis for the Southern Resident population of the Killer Whale (Orcinus orca). Centre for Biological Diversity, Tucson, Arizona. 30 p. Theriault, J.A., pers. comm. 2007. Defence Research and Development Canada. Todd, S., P. Stevick, J. Lien, F. Marques, and D. Ketten. 1996. Behavioural effects of exposure to underwater explosions in humpback whales (Megaptera novaengliae). Canadian Journal of Zoology 74: 1661-1672. Towers, JR., G.M. Ellis, and J.K.B. Ford. 2015. Photo-identification catalogue and status of the Northern Resident Killer Whale population in 2014. Canadian Technical Report of Fisheries and Aquatic Sciences 3139: iv + 75 p. Trainer, V.L. and D.G. Baden. 1999. High affinity binding of red neurotoxins to marine mammal brain. Aquatic Toxicology 46:139-148. Trites, A.W., D.E. Bain, R.M. Williams, and J.K.B. Ford. 2002. A review of short- and long-term effects of whale watching on Killer Whales in British Columbia. Pages 165-167 in the Fourth International Orca Symposium and Workshop, September 23-28, 2002. CEBC- CNRS, France. Van Bressem, M.F., K. Van Waerebeek, and J.A. Raga. 1999. A review of virus infections of cetaceans and the potential impact of Morbilliviruses, poxviruses and papillomaviruses on host population dynamics. Diseases of Aquatic Organisms 38:53-65. Van de Vijver, K.I., P.T. Hoff, K. Das, W. Van Dongen, E.L Esmans, T. Jauniaux, J. Bouquegenau, R. Blust, and W. de Coen. 2003. Perfluorinated chemicals infiltrate ocean waters: link between exposure levels and stable isotope ratios in marine mammals. Environmental Science and Technology 37: 5545-5550. Vélez-Espino, L.A., J.K.B. Ford, H.A. Araujo, G.M. Ellis, C.K. Parken, and K.C. Balcomb. 2014. Comparative demography and viability of northeastern Pacific resident killer whale populations at risk. Can. Tech. Rep. Fish. Aquat. Sci. 3084: v + 58 p. Volpe, J.P., E.B. Taylor, D.W. Rimmer, and B.W. Glickman. 2000. Natural reproduction of aquaculture escaped Atlantic salmon (Salmo salar) in a coastal British Columbia River. Conservation Biology 14: 899-903. Wade, P.R. and T. Gerrodette. 1993. Estimates of cetacean abundance and distribution in the eastern tropical Pacific. Report of the International Whaling Commission 43: 477-493. Walker, L.A., L.A. Cornell, K.D. Dahl, N.M. Czekala, C.M. Dargen, B. Joseph, A.J.W. Hsueh, and B.L. Lasley. 1988. Urinary concentrations of ovarian steroid hormone metabolites and bioactive follicle-stimulating hormone in Killer Whales (Orcinus orca) during ovarian cycles and pregnancy.Biology of Reproduction 39: 1013-1020. Waples, R. and P. Clapham. 2004. Appendix 6: Report of the working group on Killer Whales as a case study. Pages 62-73 in R.R. Reeves, W.F. Perrin, B.L. Taylor, C.S. Baker, and S.L.Mesnick (editors). Report of the Workshop on Shortcomings of Cetacean Taxonomy in Relation to Needs of Conservation and Management April 30-May 2, 2004, La Jolla, California. NOAA Technical Memorandum NOAA-NMFS-SWFSC-363. 94 pp. Available at: https://www.nwfsc.noaa.gov/assets/4/6420_09172014_185257_Waples.and.Clapham.2 004-NMFS-SWFSC-363.pdf [accessed January 2005].

272 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Ward, E.J., E.E. Holmes, and K.C. Balcomb. 2009. Quantifying the effects of prey abundance on killer whale reproduction. Journal of Applied Ecology 46(3): 632-640. WDOE (Washington Department of Ecology). 2004. Vessel entry and transit for Washington waters VEAT 2003. WDOE Publication 04-08-002. Olympia Washington. Available at: https://fortress.wa.gov/ecy/publications/publications/0408002.pdf [accessed December 2004].].].]. Weilgart, L.S. 2007. The impacts of anthropogenic ocean noise on cetaceans and implications for management. Canadian Journal of Zoology 85: 1091-1116. Whitehead, H. 2003. Sperm whales: social evolution in the ocean. University of Chicago Press, Chicago. Whitehead, H. and L. Rendell. 2004. Movements, habitat use and feeding success of cultural clans of South Pacific Sperm Whales. Journal of Animal Ecology 73: 190-196. Whitehead, H., L. Rendell, R.W. Osborne, and B. Wursig. 2004. Culture and conservation of non-humans with reference to whales and dolphins: review and new directions. Biological Conservation 120: 431-441. Wiles, G.J. 2004. Washington state status report for the Killer Whale. Washington Department of Fish and Wildlife, Olympia. 106 p. Wilkening, K.E., L.A. Barrie, and M. Engle. 2000. Trans-Pacific air pollution. Science 290: 65-66 Williams, R. 1999. Behavioural responses of Killer Whales to whale-watching: opportunistic observations and experimental approaches. M.Sc. Thesis, University of British Columbia, Vancouver, BC, Canada. Williams, R., D.E. Bain, J.K.B. Ford, and A.W. Trites. 2002b. Behavioural responses of male Killer Whales to a ‘leapfrogging’ vessel. Journal of Cetacean Research and Management 4(3): 305-310. Williams, R., C. Erbe, E. Ashe, A. Beerman, and J. Smith. 2014. Severity of killer whale behavioral responses to ship noise: A dose–response study. Marine Pollution Bulletin 79: 254–260. Williams, R., D. Lusseau, and P.S. Hammond. 2006. Estimating relative energetic costs of human disturbance to Killer Whales (Orcinus orca). Biological Conservation 133: 301- 311. Williams, R., A.W. Trites, and D.E. Bain. 2002a. Behavioural responses of Killer Whales (Orcinus orca) to whale-watching boats; opportunistic observations and experimental approaches. Journal of the Zoological Society of London 256: 255-270.

Wright, B.M., E.H. Stredulinsky, G.M. Ellis, and J.K.B. Ford. 2016. Kin-directed food sharing promotes lifetime natal philopatry of both sexes in a population of fish-eating killer whales (Orcinus orca). Animal Behaviour 115: 81-95. Ylitalo, G.M., C.O. Matkin, J. Buzitis, M.M. Krahn, L.L. Jones, T. Rowles, and J.E. Stein. 2001. Influence of life-history parameters on organochlorine concentrations in free-ranging Killer Whales (Orcinus orca) from Prince William Sound, AK. Science of the Total Environment 281: 183-203.

273 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Yurk, H., L. Barrett-Lennard, J.K.B. Ford, and C.O. Matkin. 2002. Cultural transmission within maternal lineages: vocal clans in Resident Killer Whales in southern Alaska. Animal Behavior 63: 1103-1119.

274 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Appendix A. Record of cooperation and consultation

Northern and Southern Resident Killer Whales are listed on Schedule 1 of the Species at Risk Act (SARA) and as an aquatic species are under federal jurisdiction and managed by Fisheries and Oceans Canada (DFO): 200 - 401 Burrard Street, Vancouver, BC. Southern Resident Killer Whales are a transboundary population and the United States has developed a recovery plan for Southern Resident Killer Whales as mandated under their Endangered Species Act.

To assist in the development of an initial draft of this recovery strategy, DFO brought together a diverse team of experts from various government, environmental, eco-tourism and non- governmental groups from both Canada and the United States. On the advice of the Species at Risk Coordinator at the BC Aboriginal Fisheries Commission, a letter of invitation followed up by phone calls was sent to all coastal First Nations seeking their interest in participating on the Recovery Team and/or Technical Workshop. No response was received from First Nations for inclusion on either initiative. Subsequent to the consultation process the Namgis First Nation has indicated an interest to be involved in future action planning and local implementation. A Technical Workshop was hosted in March 2004 to provide a forum for the sharing of knowledge and expertise on Killer Whales with an invited group of scientific and technical stakeholders which was invaluable in assisting the Resident Killer Whale Recovery Team to formulate an effective recovery strategy.

Public news releases announcing the Recovery Team and development of the recovery strategy and a notice of Public Consultations were sent to a distribution list of whale-related contacts provided to DFO in recent years from environmental groups, the eco-tourism sector, non- governmental organizations, government agencies and private citizens. An announcement was also placed in the Vancouver Aquarium Aquanews newsletter.

Additional input was sought through the internet (March 2005) on the draft recovery strategy and a discussion guide and feedback form were available. Responses were received from ecotourism, non-government organizations, and the Mowachaht/Muchalaht First Nation. Input from the United States National Oceanic and Atmospheric Administration and the State of Washington Department of Fish and Wildlife was received through team participation. Feedback on the recovery strategy was also received from other government agencies including: the Department of National Defence, Province of BC, SARA Secretariat, Environment and Climate Change Canada and Natural Resources Canada. An external peer review was conducted by Volker Deecke, Ph.D., University of British Columbia, and Christophe Guinet, Centre d’Etudes Biologiques de Chize, France. All feedback from both government agencies and peer reviewers was incorporated into the recovery strategy.

A proposed version of the original recovery strategy was posted on the SARA Public Registry for a 60-day public comment period, from June 21 to August 20, 2007. During this time, numerous comments were received from a wide variety of sources including government agencies, commercial and recreational fishing groups, ecotourism operators, non-governmental organizations, and private citizens. All feedback from this comment period was considered and incorporated into the recovery strategy as appropriate. Following the public comment period, the proposed Recovery Strategy was revised by DFO in order to address public comments and to reflect the responsibilities of the competent Minister.

Minor amendments to the recovery strategy were made in 2011 to provide additional clarification regarding critical habitat for Northern and Southern Resident Killer Whales.

275 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Consultations were not held on the amendments as significant changes were not made to the recovery strategy.

In 2018, the recovery strategy was amended again to include identification of additional critical habitat for these populations and to provide minor updates to background and species information. A 30-day online external review on the draft amended recovery strategy was held June 12 – July 11, 2018. The external review was targeted for those potentially affected to provide feedback on the draft amended recovery strategy prior to public consultation. Numerous comments were received on the draft amended recovery strategy. 670 comments were submitted via an online feedback form on the consultation webpage set up by DFO’s Regional SARA Program and 22 letters were received by email. Feedback was received from numerous groups and organizations inclusive of Wildlife Management Boards, Indigenous groups, the province of British Columbia, local government, commercial and recreational fishers, ecotourism operators, non-governmental organizations, other government agencies, and private citizens.

Additional Indigenous, stakeholder, and public input was sought through the publication of the proposed document on the Species at Risk Public Registry for a 60-day national online public comment period from September 4 – November 3, 2018. During this 60-day period, meetings with Indigenous groups and Wildlife Management Boards were held as requested, as well as two webinars (one for Indigenous participants and one for members of the public), and two community engagement sessions. Input was received from a wide variety of groups, individuals, and organizations, including Wildlife Management Boards, Indigenous groups, stakeholders, non-governmental organizations, and local government. In total, 162 individual comments were received, as well as 38 letters, two petitions and one form letter campaign. All comments received during the external review and public comment period were considered and resulted in minor changes to the document.

276 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Appendix B. Recovery team members

The following individuals composed the Recovery Team for the development of the 2008 recovery strategy. Marilyn Joyce Co-Chair: Resident Killer Whale Recovery Team Fisheries & Oceans Canada, Fisheries Management Branch Pacific Region, 200-401 Burrard Street, Vancouver, B.C., V6C 3S4, phone: 604-666-9965, email: [email protected] Lance Barrett-Lennard Co-Chair: Resident Killer Whale Recovery Team Vancouver Aquarium Marine Service Center Stanley Park, Vancouver, BC V6B 3X8, phone: 604-659-3428, email: [email protected]

David Bain Friday Harbor Laboratories, University of Washington, WA Ken Balcomb Centre for Whale Research, WA Jim Borrowman North Island Whale Watching Community, BC John Durban National Marine Fisheries Service, Alaska Fisheries Science Centre, National Marine Mammal Laboratory, WA Graeme Ellis Fisheries & Oceans Canada, Science Branch, Conservation Biology Section, BC John Ford Fisheries & Oceans Canada, Science Branch, Conservation Biology Section, BC Christine Garrett Environment and Climate Change Canada, Environmental Protection Branch, Commercial Chemicals Division , BC Anna Hall Whale Watch Operators Association North West, BC Steve Jeffries Washington Department of Fish and Wildlife, Marine Mammal Investigations, WA Linda Jones National Marine Fisheries Service, Northwest Fisheries Science Centre, National Marine Mammal Laboratory, WA Brent Norberg National Marine Fisheries Service, Protected Resources Division, WA Peter Olesiuk Fisheries & Oceans Canada, Science Branch, Conservation Biology Section, BC Rich Osborne The Whale Museum, WA Rob Paynter Ministry of Sustainable Resource Management, BC Brian Reader Western Canada Service Centre, Parks Canada Agency, BC Peter Ross Fisheries and Oceans Canada, Marine Environmental Quality Section, BC Paul Spong Orcalab, Hanson Island, BC Andrew Trites Marine Mammal Research Unit, Fisheries Centre, University of British Columbia, BC Scott Wallace (Alternate) Marine Conservation Caucus, Raincoast Conservation Society Sierra Club of Canada, B.C. Chapter, BC

277 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Gary Wiles (Alternate) Washington Department of Fish and Wildlife, Marine Mammal Investigations, WA Rob Williams Marine Conservation Caucus, Raincoast Conservation Society, BC Brian Riddell Fisheries & Oceans Canada, Science Branch, Salmon and Freshwater Ecosystems, BC

Resource Personnel:

Paul Cottrell Fisheries & Oceans Canada, A/SARA First Nations Coordinator, Treaty & Aboriginal Policy Branch, BC Carole Eros Fisheries & Oceans Canada, Species at Risk Recovery Planning Coordinator, Resource Management Pacific Region, BC Annely Greene Fisheries & Oceans Canada, Marine Mammal Program Manager, Resource Management Pacific Region, BC Kathy Heise Department of Zoology, University of British Columbia, BC Lara Sloan Fisheries & Oceans Canada, Communications Officer, Fisheries Management Pacific Region, BC

278 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Appendix C. Contaminant acronyms

APEs: Alkylphenol ethoxylates DBT: Dibutyltin DDT: Dichlorodiphenyl trichloroethane PAHs: Persistent aromatic hydrocarbons PBDEs: Polybrominated diphenylethers PBDTs: Polybrominated trienylethers PCBs: Polychlorinated biphenyls PCDDs: Dioxins, polychlorinated dibenzo-p-dioxins PCDFs Polychlorinated dibenzofurans PCNs: Polychlorinated napthalenes PCPs: Polychlorinated paraffins PCTs: Polychlorinated terphenyls SPFOs: Perfluoro-octane sulfonates POPs: Persistent organic pollutants TBT: Tributyltin

279 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Appendix D. Description of critical habitat

Southern Resident Killer Whale critical habitat boundaries for transboundary waters of southern Georgia, Haro, and Juan de Fuca Straits. Described clockwise from the western boundary - all Latitudes are Decimal Degrees North; all Longitudes are Decimal Degrees West.

Start and end coordinates Latitude Latitude Longitude Longitude Point description Deg Min Deg Min 1 48 29.68 124 44.31 Western boundary 2 48 40.02 124 50.68 3 48 21.30 123 44.32 Excluding waters north of the line joining (Sooke Inlet) 4 48 20.33 123 42.90 5 Excluding waters north of the line joining (Royal Roads, 48 24.25 123 28.97 6 Esquimalt Hbr, Victoria Hbr) 48 24.57 123 22.61 7 Excluding waters west of the line joining (Cordova Channel 48 29.69 123 18.61 8 and Sidney Channel) 48 36.12 123 18.51 9 Excluding waters west of the line joining (western half of 48 37.04 123 18.49 10 Miners Channel and the waters west of Gooch Island) 48 39.70 123 17.72 11 Excluding waters west of the line joining (western half of 48 39.88 123 17.68 12 Prevost Channel and Moresby Passage) 48 42.96 123 19.63 13 Excluding waters west of the line joining (western portion of 48 43.34 123 19.88 Swanson Channel between Moresby Island and Prevost 14 Island) 48 48.86 123 22.70 15 Excluding waters west of the line joining (western portion of 48 50.66 123 23.33 Trincomali Channel between Prevost Island and Parker 16 Island) 48 52.61 123 23.92 17 Excluding waters west of the line joining (western portion of 48 52.85 123 23.92 Trincomali Channel between Parker Island and Galiano 18 Island) 48 53.08 123 23.76 19 48 54.28 123 20.67 20 48 55.39 123 21.98 Excluding waters west of the line joining (western portion of 21 49 0.00 123 18.88 southern Strait of Georgia) 22 49 10.39 123 22.82 23 49 13.58 123 21.97 24 49 13.58 123 21.97 25 Excluding waters north of the line joining (portion of southern 49 14.00 123 21.09 26 Strait of Georgia) 49 14.18 123 19.22 27 49 13.79 123 17.21 28 49 13.79 123 17.21 29 49 12.87 123 15.75 Excluding waters north and east of the line joining (portion of 30 49 9.01 123 16.48 southern Strait of Georgia) 31 49 3.39 123 9.24 32 49 3.47 123 8.48 And bounded on the east and south by Point Roberts and

the United States Border

280 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Northern Resident Killer Whale critical habitat boundaries for Johnstone and southeastern Queen Charlotte Straits. Described clockwise from the western boundary - all Latitudes are Decimal Degrees North; all Longitudes are Decimal Degrees West.

Start and end coordinates Latitude Latitude Longitude Longitude Point Description Deg Min Deg Min 1 50 36.98 127 11.00 Western boundary (Vancouver Island to Numas Island) 2 50 46.24 127 6.76 3 50 46.27 127 5.26 Northern boundary (Numas Island to Broughton Island) 4 50 46.41 126 48.27 5 Northern boundary (Broughton Island to Screen Island / Eden 50 46.13 126 47.30 6 Island) 50 44.95 126 43.55 7 boundary line running from Eden Island to Crib Island 50 44.79 126 43.22 (including waters of Queen Charlotte Strait and excluding 8 waters of Trainer Passage) 50 43.67 126 42.73 9 boundary line running from Crib Island to House Ilet (including 50 43.33 126 42.58 waters of Queen Charlotte Strait and excluding waters of 10 Arrow and Spring Passages) 50 40.16 126 41.21 11 boundary line running from House Ilet to Swanson Island 50 40.16 126 41.21 (including waters of Queen Charlotte Strait and excluding 12 waters of Knight Inlet) 50 37.75 126 43.86 13 boundary line running from Swanson Island to Compton 50 36.06 126 41.77 Island (including waters of Blackfish Sound excluding waters 14 of West Passage) 50 35.84 126 41.42 15 boundary line running from Compton Island to Harbledown 50 35.50 126 40.86 Island (including waters of Blackfish Sound excluding waters 16 of Whitebeach Passage) 50 35.38 126 40.68 17 boundary line running from Harbledown Island to Parson 50 35.19 126 40.93 Island (including waters of Blackfish Sound excluding waters 18 of Parson Bay) 50 34.43 126 40.73 19 boundary line running from Parson Island to West Cracroft 50 33.65 126 39.95 Island (including waters of Blackfish Sound excluding waters 20 of Baronet Passage) 50 32.98 126 39.73 Waters of western Johnstone Strait bounded on the north by West Cracroft Island, the mainland, Hardwicke Island and West Thurlow Island with no exclusions except: 24 boundary line running from West Cracroft Island to the 50 31.32 126 20.35 mainland (including waters of western Johnstone Strait 25 excluding waters of Havannah Channel) 50 31.09 126 17.05 26 boundary line running from the mainland to Hardwicke Island 50 28.46 126 2.54 (including waters of western Johnstone Strait excluding 27 waters of Sunderland Channel) 50 26.57 125 57.94 28 boundary line running from Hardwicke Island to Eden Point 50 24.58 125 48.29 on West Thurlow Island (including waters of western 29 Johnstone Strait excluding waters of Chancellor Channel) 50 23.91 125 47.38 30 boundary line running from Eden Point to Tyee Point on West 50 23.91 125 47.38 Thurlow Island (including waters of western Johnstone Strait 31 excluding waters of Vere Cove) 50 23.26 125 47.06 32 Eastern boundary line running from West Thurlow Island 50 23.42 125 34.39 (including waters of western Johnstone Strait excluding 33 50 21.88 125 34.23 waters of eastern Johnstone Strait and Mayne Passage) Waters of western Johnstone Strait bounded on the south by

Vancouver Island - no exclusions except: 35 boundary line running from Graveyard Point to Kelsey Bay 50 23.45 125 56.71 Harbour on Vancouver Island (including waters of western 36 Johnstone Strait excluding waters of Salmon Bay) 50 23.80 125 57.62

281 Recovery Strategy for Northern and Southern Resident Killer Whales 2018

Northern and Southern Resident Killer Whale critical habitat boundaries off Southwestern Vancouver Island. Described counter-clockwise from the northern boundary - all Latitudes are Decimal Degrees North; all Longitudes are Decimal Degrees West.

Start and end coordinates Point Description Latitude Latitude Longitude Longitude Deg Min Deg Min 1 Northern Boundary (Vancouver Island running 48 59.7 -125 40.15 southwest offshore) 2 48 41.72 -126 17.88 3 Offshore Boundary 48 13.95 -125 44.61 4 Waters adjacent the U.S.A. Border 48 29.72 -124 44.32 5 Waters adjacent Southern Resident Killer Whale 48 40.04 -124 50.66 critical habitat in transboundary waters of southern Georgia, Haro, and Juan de Fuca Straits 6 And bounded by Vancouver Island to the Northwest boundary 7 Excluding waters north of the line joining (Nitinat Inlet) 48 40.05 -124 50.99 8 48 40.13 -124 51.3 9 Excluding waters northeast of the line joining Cape 48 55.22 -125 32.391 Beale and Amphitrite Point (Barkley Sound) 10 48 47.174 -125 13.039

Northern Resident Killer Whale critical habitat boundaries in western Dixon Entrance. Described clockwise from the western boundary - all Latitudes are Decimal Degrees North; all Longitudes are Decimal Degrees West. Start and end coordinates Latitude Latitude Longitude Longitude Point Description Deg Min Deg Min 1 54 15.38 -133 3.5 Western Boundary (Langara Island Northward) 2 54 15.99 -133 3.5 3 Northern Boundary 54 16.05 -131 40.45 4 Eastern Boundary 54 9.13 -131 40.43 5 Excluding waters south of line (McIntyre Bay) 54 5.491 -132 15.97 Bounded by Graham Island on the Southern 54 11.07 -133 1.55 6 Boundary to Northward to Langara Island, excluding waters west 54 11.43 -133 0.75 7 of the line Bounded on the western Boundary by the eastern 8 side of Langara Island up to Langara Light 9 Excluding waters south of line (Virago Sound, Naden 54 5.86 -132 26.26 Harbour) 10 54 5.57 -132 34.3

282

Canadian Science Advisory Secretariat Pacific Region Science Response 2018/035

SCIENCE INFORMATION TO SUPPORT CONSULTATIONS ON BC CHINOOK SALMON FISHERY MANAGEMENT MEASURES IN 2018

Context The 2018 draft Integrated Fisheries Management Plans (IFMP) for Pacific Salmon include a number of proposed fisheries management measures for Chinook Salmon (Oncorhynchus tshawytscha) in 2018. Fisheries Management has requested that DFO Science provide information (trends in abundance, productivity and current exploitation) for key Chinook Salmon management units to support consultations on potential additional BC Chinook Salmon fishery management measures in 2018.

This Science Response (SR) represents the best available science information on Chinook Salmon at present compiled on a short timeline. Therefore, the data and interpretations presented here are subject to change as new information becomes available.

This Science Response Report results from the Science Response Process on April 11, 2018 on Science information to support consultations on BC Chinook Salmon fishery management measures in 2018. Background Large-scale patterns of environmental change and increased environmental variability have been associated with broad declines in productivity1 of Chinook Salmon across their range in recent decades. Potential effects of recent events, such as the persistence of the warm ocean water ‘blob’ which formed in the North Pacific in 2014 and moved onshore in 2015, and El Niño conditions in early 2016, have lowered expectations for returns of Chinook Salmon in 2018 (PFMC 2018).

Chinook Salmon spawn in many river systems covering a broad geographic range from California to Alaska along the west coast of North America. Given this broad geographic distribution, encompassing diverse spawning habitats for adults and rearing habitats for juveniles, and the wide variation in early life history, age of maturation, ocean distribution, return timing, and other characteristics, typically only some populations will show declines in response to acute, single-season environmental pressures. These wide-ranging traits also complicate evaluation of the influence of factors affecting Chinook Salmon productivity, abundance and survival. For example, increasing frequency and intensity of El Niño events through the 1990s have resulted in increasing variance in the North Pacific Gyre Oscillation and have been associated with declines in Chinook Salmon populations coast-wide and increased survival rate

1 Productivity is the intrinsic rate of growth of a population, estimated from the observed relationship between spawners and adult recruits over time (also referred to as “recruits per spawner”).

May 2018 283 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018 synchrony (Kilduff et al. 2014; Mantua 2015). Varying responses at the local population level have been observed during the same time period.

Declines in productivity and abundance of many southern Chinook Salmon stocks were observed during consecutive large El Niño events in the early and late 1990s. Stock groups such as West Coast Vancouver Island and Strait of Georgia experienced dramatic declines in marine survival rates and resulting productivity. Other stocks such as the Fraser Summer 41 management unit (e.g., South Thompson ocean-type summer run timing Chinook Salmon) were less affected by these changes, but these stocks as well as others have recently shown declines in abundance, productivity and survival rate. The most recent Wild Salmon Policy (WSP) integrated biological status assessment of Southern BC Chinook Salmon identified 11 conservation units (CUs) as ‘red’ (i.e. spawning abundance is likely below the lower biological benchmark) out of 15 CUs for which consensus was reached on an integrated status designation; an additional nine CUs were designated as data deficient and status could not be evaluated for 11 CUs (Table 1; DFO 2016).

Transboundary and Northern BC Chinook Salmon stocks (e.g., Alsek, Taku, Stikine, Skeena and Nass rivers), which in the 1990s and 2000s appeared to maintain a higher productivity, are showing more recent declines in abundance and productivity. Recent declines are also apparent for the Fraser Summer 41 (South Thompson) Chinook Salmon and other Fraser River stocks, especially the stream-type spring and summer run timed stocks. In contrast, Southern BC coastal stocks, which had the greatest initial decline in productivity and remained at low levels, have recently exhibited some increases in escapement (Figure 1), particularly along the east coast of Vancouver Island. Note that the degree to which these increases can be associated with increases in productivity is uncertain at present. In contrast, marine survival rates of east coast Vancouver Island stocks have remained consistently below their time series average (Figure 2a,b). Along the west coast of Vancouver Island, recent marine survival rates from Robertson Creek Hatchery appear to be near average historic levels, while low abundance of local wild populations, such as those in Clayoquot Sound, remains a concern in southwest Vancouver Island.

Key observational data in British Columbia are derived from ‘indicator’ stocks distributed throughout BC. Spawning abundance is estimated for each indicator stock using methods ranging from high precision fence counts to lower precision escapement enumeration methods. Coded Wire Tag (CWT) indicator stocks, generally associated with hatcheries, provide information such as marine survival, fishery exploitation rates, and ocean distribution information. These data are tracked by the Pacific Salmon Commission and results are accessible through the publications of the Chinook Technical Committee. The key Chinook Salmon management units under consideration and associated indicator stock data are summarized in Table 1. The information from these indicator stocks shows regional variation in escapement abundance and marine survival rate trends (Figure 1 and Figure 2a,b). At finer spatial scales, local habitat and ecosystem factors may explain some variations in abundance. In some cases, such as the Cowichan River Chinook Salmon stock, watershed and habitat restoration may be important factors in recent increased returns.

Dorner et al. (2017) associated the broad pattern of declines in Chinook Salmon productivity, from Alaska to Oregon, with unfavourable large-scale climatic change in the North Pacific Gyre Oscillation and the North Pacific Current, as well as increased frequency of large scale events such as El Niño, and in 2014-15, the persistence of warm ocean waters in ‘the blob’. Other researchers such as Ohlberger et al. (2018) suggest that the biological mechanisms behind the decline in productivity also include changes in population demographics, such as younger age-

2 284 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018 at-maturity, reduced size-at-age, and reduced fecundity of female spawners. Some of these demographic effects are now being observed in BC Chinook Salmon populations (Table 2, Figure 3a,b). Selective exploitation of large Chinook Salmon is likely a contributing factor to the decline in body size (Ohlgerger et al. 2018) and other demographic changes, as is predation by seals, sea lions, killer whales, and salmon sharks (Ford et al. 2009; Trites and Rosen 2018; Chasco et al. 2016; Nagasawa 1998). In addition, degradation of freshwater spawning and rearing habitat may contribute to the longer term declines in productivity observed in many BC management units (summarized in Riddell et al. 2013).

Sustainable exploitation rates (EMSY) are directly related to productivity; when productivity declines, fishery exploitation should be reduced2. Since the early 1980s, two of five southern B.C. indicator stocks with productivity estimates have exhibited declines in productivity exceeding 40%, while productivity for a third indicator stock shows declines of approximately 25% and the remaining two are relatively stable (Riddell et al. 2013). The associated reduction in sustainable exploitation rate depends on the initial productivity of the stock. Marine area fishery catch and exploitation were reduced over time, starting with the first Pacific Salmon Treaty in 1985. In southern BC, total Chinook Salmon catch was reduced by 78% from the early 1980s (Table 3). In northern BC, total marine Chinook Salmon catch was reduced by about 47% from the early 1980s (Table 4). Resulting annual exploitation rates were reduced by an average of 44% for BC CWT indicator stocks (Table 5 and Figure 4) because stocks were considered to be overexploited at the time. Formal recognition of the link between productivity and sustainable exploitation supports the need for re-evaluation of prior ER estimates that were thought to be conservative (Hilborn & Walters, 1992). Dorner et al. (2017) suggest that, for some BC stocks, there have been declining trends in productivity ranging from about 15 to 66% over the available time series (brood years 1979-2008; Table 6). Based on these trends in productivity, further reductions in exploitation may also be warranted. Analysis and Response Fishery Managers requested DFO Science provide and organize available data and other information to address the questions outlined below. These responses will facilitate consultation with First Nations, Industry and other stakeholders in the development of additional fishery management measures that may be required to address declines in Chinook Salmon stock productivity.

Q1. Provide information to determine which stocks require a reduction in fishery exploitation.

Criteria (based on information available at this time) that may be used to determine which stocks may require further management measures to adjust fishery impacts include:

• Recent average exploitation rates relative to estimates of sustainable exploitation (EMSY) given current stock productivity (Table 6); • Level of recent escapement (and forecasts for WCVI stocks only) relative to escapement goals (Table 7); • Evidence of recent declines in marine survival rate (Figure 2);

2 EMSY, or sustainable exploitation rate is derived from the ‘Ricker a’ parameter, an estimate of productivity, by numerical approximation (Hilborn and Walters 1992).

3 285 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

• Identification of other fishery related impacts, such as selective fishing practices, that may be contributing to declines in stock productivity; and • WSP integrated status assessments. A significant issue for these preliminary analyses is that data quality and amount varies across management units. For example, the WSP integrated biological status assessment for southern BC Chinook resulted in status assessments for 15 stocks but identified 11 data deficient CUs for which integrated status assessment could not be determined. There are also inherent sources of uncertainty associated with the available data and with the estimation of management parameters (such as EMSY) but it is the best information currently available to inform the current request for Science advice. Examples of inherent uncertainty include the CWT expansions used in sport and First Nations fisheries, and estimates of total catch in some fisheries. Further work is required to develop and evaluate stock assessment methods that can be applied to more data limited Chinook Salmon stocks. Science is currently developing these methods and more complete information to inform data limited assessments and risk-appropriate management responses will be available through future work (e.g., the Fisheries Management Framework Initiative). Forecasts are provided for a limited number of stocks, but their performance has varied historically and the methods used have not been fully reviewed by Science.

Q2. What tools and information can Science provide to inform trade-offs associated with a range of potential reductions in fishery exploitation rates?

Reductions in fishery exploitation rates to achieve stock rebuilding objectives and fishery objectives inherently involve management choices such as rebuilding times and risk tolerance (i.e., the probability of achieving those objectives). Co-management processes, such as the Southern BC Chinook Initiative, are recommended as appropriate venues to conduct such management strategy evaluations due to their inclusivity and broad representation. While the Southern BC Chinook Initiative is underway, results from the initiative will not be available in a suitable time frame to assist with current management objectives. Information that may be useful in considering reductions in fisheries includes run timing, CWT- based distribution of total mortality, creel survey interview data and biological sampling, the iREC survey of licence holders, DNA stock composition, and commercial and test fishery data. For northern stocks, this information is available through the results of a special joint initiative in 2018 involving DFO biologists and biologists from the Alaska Department of Fish and Game (available from I. Winther, DFO by request). In southern BC, recreational fishery data summaries from creel surveys and iREC surveys are available from W. Luedke, DFO by request. Summaries of the CWT data and associated mortality distribution tables current to 2017 are available upon request (G. Brown or C. Parken, DFO). These types of data have been used in simulations or retrospective analyses to explore a range of fishery-specific reductions intended to reduce fishery impacts and to increase passage of fish to spawning grounds (e.g., Starr and Argue, unpublished working paper3). These data products can identify benefits and costs associated with a range of fishery reductions and management strategies. Simulations in the form of retrospective scenarios of assumed Canadian and US fishery reductions using the PSC Coast-Wide Chinook Model have recently been conducted to better understand the likely range of stock-specific increases in abundance and also, the impact

3 Starr, P.J., and Argue, A.W. 1991. Evaluation Framework for Assessing 1989 Strait of Georgia Sport Fishing Regulation Changes. Pacific Stock Assessment Review Committee. Working paper 91-3.

4 286 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018 to fishery catches, The development of scenarios have been informed by the use of stock- and fishery-specific CWT-based estimates of exploitation rates. Further simulation work is possible and would be best guided by management input in the form of development of target objectives for evaluation, and the identification of potential management strategies to achieve those objectives. Science, working with other DFO sectors and through various joint technical processes involving First Nations and stakeholders, is currently completing work that will provide more comprehensive advice to adequately inform the decision-making context. This work includes developing stock assessment methods that can be applied for more data limited situations, developing robust methods for estimating sustainable exploitation rates, and developing evaluation tools that can be used to inform management trade-offs when setting fishery and stock objectives for rebuilding. There is also a technical review underway to evaluate management actions implemented in 2012 to reduce fishery impacts on Fraser River Chinook Salmon. As this work is completed, the information that Science can provide to managers will be more comprehensive and robust. Q3. What information can Science provide to inform development of management measures if it is determined reductions are required? Declines in productivity as described in the Background section may warrant either reduction in exploitation rates and/or measures to reduce fishery-related impacts that may contribute to negative demographic changes in populations (e.g., harvest practices that selectively remove older and larger fish). Increased exploitation of the oldest age classes have been observed in North Coast stocks, West Coast Vancouver Island stocks but not the lower Strait of Georgia or Fraser River stocks. Once the stocks of concern and target levels for potential reductions are identified, more specific input can be provided by Science to inform development of specific management measures. Methods used to assess proposed fishery measures on a by-fishery basis will depend on how proposed reductions are implemented (e.g., reductions in total allowable catch or fishing effort and area closures in times when stocks of concern are prevalent can both be informed by analyses of CWT indicator stock distributions in fisheries; while bag limits, size limits, and other gear restrictions (e.g., net mesh size) may help to mitigate demographic changes for some stocks. Science can use data including historical fishery impacts, stock distribution and timing, size at catch and fishing effort to model expected reductions in fishery impacts. Information, data availability and data quality will range from high to low across Chinook stocks in British Columbia. The use of proxy data (i.e., data from a stock considered to exhibit a similar life history pattern and ocean distribution is ‘borrowed’ for a stock with missing information) is a common approach in data limited cases. Limitations of available data are particularly important in consideration of the scale of reductions that can be modeled. Finer-scale or incremental reductions in fisheries are best supported with good quality data. Data of sufficient quality and resolution are available for some CWT stocks but data are more limited for others.

Q4. What are the potential metrics/indicators that could be used to assess whether or not objectives have been met? Provide commentary on strengths and weaknesses of proposed assessment methods.

Metrics/indicators that could be used to assess whether or not objectives have been met should be similar to the criteria used to set targets for reduction. That is, for the management units in which management actions are taken, performance metrics could include:

5 287 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

• A reduction in observed exploitation rate to a level below EMSY (which will provide a buffer to better account for the uncertainty associated with estimation of EMSY); • An increase in escapement of indicator stocks within the management unit (i.e., observed rebuilding to a level above an established benchmark, such as the WSP absolute abundance metric Sgen); • An observed reduction in size-selective fishery impacts; and • Observed reversal of declining trends in escapement (e.g., improvement in established metrics from lower WSP zone, red status to amber or green). In all cases, the sources of uncertainty associated with potential metrics and data deficiency should be considered and targets set accordingly. The ability to assess the achievement of specific reduction targets post-season on a by-fishery basis is dependent on catch monitoring and sampling programs conducted during the fishing season. The choice of performance metrics/indicators identified on a by-fishery basis that are inconsistent with current stock assessment and catch monitoring frameworks (or highly sensitive to the uncertainty of the available data) may require additional monitoring and sampling programs. Some escapement programs produce relatively imprecise estimates of spawning abundance. Finally, detecting measurable improvements associated with fishery actions taken in 2018 may not be possible given inter-annual variation in environmental conditions that influence marine survival rate and stock abundance. Key Observations • Chinook Salmon productivity is estimated to have declined from 25-40% since the early 1980s across many BC indicator stocks. According to stock assessment principles, sustainable exploitation (EMSY) is directly related to productivity; when productivity declines, fishery exploitation should be reduced. • Reliable estimates of productivity rely on complete brood year information (which is only available once the oldest age for the cohort has spawned and coincidentally once all fisheries have occurred on the cohort); presently, current productivity estimates are only available up to the 2012-2013 brood years and do not reflect any recent changes in productivity. The uncertainty introduced by this information lag, along with low escapements and recent declines in calendar year exploitation rates may explain why additional reductions in exploitation are being considered despite recent exploitation rates falling below EMSY. • Factors potentially affecting the productivity of recent brood years (i.e., that are not reflected in the available EMSY estimates) include anomalous ocean conditions such as “the blob” in 2014-15, an El Niño in the first half of 2016 and freshwater habitat issues relating to drought conditions in the southern interior B.C. (2015-2016). Further, anecdotal reports suggest that there was an acute decrease in abundance of juvenile Chinook Salmon entering the ocean in the summer of 2017 (observed in standardized surveys along the Washington coast and in Alaska; L. Weitkamp, pers. comm., January 2018), which suggests adult returns for these brood years may also be low (roughly corresponding to returns in 2018 through to 2021). • Patterns in Chinook Salmon productivity have also become more synchronous in recent years, similar to results reported for other species of Pacific salmon. Such recent changes may reduce the resilience of the species to effects of climate change and habitat modification (Dorner et al. 2017).

6 288 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

• As noted previously, a number of key uncertainties are associated with the data presented here. Due to the wide-ranging age-at-return of many Chinook populations, complete brood year information can take up to six years to accrue and modelled estimates are often used in the interim. Additionally, the productivity of hatchery populations may differ from associated natural populations, but the available data does not provide a means of differentiating between hatchery and natural populations. This limits the ability to make inferences about natural productivity which is of greatest concern. • Contrary to the basin-scale pattern of declines in productivity, increases in escapement have been observed for some Vancouver Island stocks; at present, the specific natural and/or human-caused mechanism(s) leading to these increases have not been isolated. Conclusions 1. Declining productivity is still a concern for many Chinook Salmon stocks and it is unclear whether current estimates of sustainable exploitation are appropriate for current productivity levels. 2. Given the variability and time lag in the existing data, detecting changes in productivity will take several years. General trends can be identified from broader consideration of changes in recent escapement and calendar year exploitation rates along with brood year information, despite the uncertainties inherent in these data sources. 3. While there is an attempt to identify sources of uncertainty in this preliminary response to inform the broader decision-making process, these uncertainties have not been presented with sufficient detail to fully understand their impact on local scale decision-making processes. 4. A range of initiatives are currently underway that will improve the ability to provide science advice to support management decision making, such as the Southern BC Chinook Initiative. Additionally, information that may be useful in considering reductions to specific stocks that are currently available, include (but are not limited to) run timing, CWT-based distribution of total mortality, creel survey interview data and biological sampling, the iREC survey of licence holders, DNA stock composition, commercial and test fishery data, recent simulation work with the PSC Coast-wide Chinook Model, and additional information from DFO members of the CTC. The sources of data available vary by stock and area. 5. Science, working with other DFO sectors and through various joint technical processes involving First Nations and stakeholders, is currently completing work that will provide more comprehensive tools and methods to inform the decision-making context. Contributors Contributor Affiliation Gayle Brown DFO Science, Pacific Region Diana Dobson DFO Science, Pacific Region Chuck Parken DFO Science, Pacific Region Mary Thiess DFO Science, Pacific Region Ivan Winther DFO Science, Pacific Region Wilf Luedke DFO Science, Pacific Region Dawn Lewis DFO Science, Pacific Region Antonio Velez-Espino DFO Science, Pacific Region

7 289 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Contributor Affiliation Michael Folkes DFO Science, Pacific Region John Holmes DFO Science, Pacific Region Lesley MacDougall DFO Science, Pacific Region

Approved by Carmel Lowe Regional Director Science Branch, Pacific Region Fisheries and Oceans Canada April 16, 2018 Sources of information Chasco, B., Kaplan, I.C., Thomas, A., Acevdeo-Gutierrez, A., Noren, D., Ford, M.J., Hanson, M.B., Scordino, J., Jeffries, S., Pearson, S., Marshall, K.N., Ward, E.J. 2017. Estimates of Chinook salmon consumption in Washington State inland waters by four marine mammal predators from 1970 to 2015. Can. J. Fish. Aquat. Sci. 74: 1173–1194. CTC (Chinook Technical Committee). 2017a. Annual Report of Catch and Escapement for 2016. Pacific Salmon Commission. TCCHINOOK (17)-2. CTC (Chinook Technical Committee). 2017b. 2016 Exploitation Rate Analysis and Model Calibration Supplement – Data Notebook. Pacific Salmon Commission. TCCHINOOK (17)-1. ( a full report has not been completed since 2014 TCCHINOOK (15)-1 V1 and V2) DFO. 2016. Integrated Biological Status of Southern British Columbia Chinook Salmon (Oncorhynchus tshawytscha) Under the Wild Salmon Policy. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2016/042. Dorner, B., Catalano, M.J., Peterman, R.M. 2017. Spatial and temporal patterns of covariation in productivity of Chinook Salmon populations of the north eastern Pacific Ocean. Can. J. Fish. Aquat. Sci. 00: 1-14. Ford, J.K.B, Wright, B.M., Ellis, G.M., and Candy, J.R. 2010. Chinook salmon predation by resident killer whales: seasonal and regional selectivity, stock identity of prey, and consumption rates. DFO Can. Sci. Advis. Sec. Res. Doc. 2009/101. iv + 43 p. Hilborn, R. and C. J. Walters. 1992. Quantitative Fisheries Stock Assessment. Chapman and Hall: New York. Kilduff, D.P., Botsford, L.W., and Teo, S.L.H. 2014. Spatial and temporal covariability in early ocean survival of Chinook Salmon (Oncorhynchus tshawytscha) along the west coast of North America. ICES J. Mar. Sci. 71(7): 1671-1682. Mantua, N.J. 2015. Shifting patterns in Pacific climate, West Coast salmon survival rates, and increased volatility in ecosystem services. PNAS September 1, 2015. 112(35): 10823- 10824; published ahead of print August 24, 2015. (Accessed April 17, 2018). Nagasawa, K. 1998. Predation by salmon sharks (Lamna ditropis) on Pacific salmon (Oncorhynchus spp.) in the North Pacific Ocean. N. Pac. Anadr. Fish Comm. Bull. No.1: 419-433

8 290 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Ohlberger, J., Ward, E.J., Schindler, D.E., Lewis, B. 2018. Demographic changes in Chinook Salmon across the Northeast Pacific Ocean. Fish and Fisheries, 2018; DOI: 10.1111/faf.12272. Parken, C.K., McNicol, R.E., and Irvine, J.R. 2006. Habitat-based methods to estimate escapement goals for data limited Chinook salmon stocks in British Columbia, 2004. DFO Can. Sci. Advis. Sec. Res. Doc. 2006/083. vii + 67 p. PFMC (Pacific Fishery Management Council). 2018. Review of 2017 Ocean Salmon Fisheries: Stock Assessment and Fishery Evaluation Document for the Pacific Coast Salmon Fishery Management Plan. (Document prepared for the Council and its advisory entities.) Pacific Fishery Management Council, 7700 NE Ambassador Place, Suite 101, Portland, Oregon 97220-1384. Riddell. B., M. Bradford, R. Carmichael, D. Hankin, R. Peterman and A. Wertheimer. 2013. Assessment of status and factors for decline of southern BC Chinook Salmon: Independent Panel’s Report. Prepared with the assistance of D.R. Marmorek and A.W. Hall, ESSA Technologies Ltd., Vancouver, B.C. for Fisheries and Oceans Canada (Vancouver, BC) and Fraser River Aboriginal Secretariat (Merritt, BC). xxix + 165 pp. + Appendices. Trites, AW and Rosen, DAS (eds). 2018. Availability of Prey for Southern Resident Killer Whales. Technical Workshop Proceedings. November 15–17, 2017. Marine Mammal Research Unit, Institute for the Oceans and Fisheries, University of British Columbia, Vancouver, B.C., 64p. Wood, C. C. 1987a. Predation of juvenile Pacific salmon by the common merganser (Mergus merganser) on eastern Vancouver Island. I: Predation during the seaward migration. Can, 8 . Fish. Aquat. Sci. 44:941 -949. Wood, C. C. 1987b. Predation of juvenile Pacific salmon by the common merganser (Mergus merganser) on eastern Vancouver Island. II: Predation of stream-resident juvenile salmon by rnerganser broods. Can. J. Fish. Aquat. Sci. 44: 950-959.

9 291 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Tables and Figures Table 1. Key characteristics of data used to inform the analysis for Transboundary and BC Chinook management units.

Life Chinook History Escapement Management Unit Type CWT Indicators CTC Model Stock Indicators Associated Conservation Units4 CK-60: Stikine_early Taku (TAK), Stikine CK-61: Stikine_late Transboundary Taku, Stikine (STI) CK-63: Taku_early Transboundary Stream CK-64: Taku_mid CK-65: Taku_late Alsek Alsek CK-67: Alsek CK-46: Ecstall CK-48: Lower Skeena CK-49: Kalum_early CK-50: Kalum_late Nass, Skeena, CK-51: Lakelse North Coast Stream Kitsumkalum (KLM) Northern BC Kitsumkalum CK-53: Middle Skeena-large lakes CK-54: Middle Skeena-mainstem tribs, CK-55: Upper Bulkley River CK-56: Upper Skeena CK-57: Portland Sd-Observatory Inlet- Lower Nass Central BC Ocean Atnarko (ATN) Central BC Atnarko CK-39: Bella Coola-Bentinck Quinsam Hatchery (QUI) CK-28: Southern Mainland-Southern Phillips River Fjords Upper Georgia (PHI)5 Upper Georgia Strait Aggregate index CK-29: East Vancouver Island-North Ocean Strait Puntledge Hatchery (PPS) Middle Georgia Strait CK-27: East Vancouver Island-Qualicum Big Qualicum & Puntledge-fall timing (BQR)

4 Conservation Units appearing in grey italics have not yet undergone a Wild Salmon Policy integrated status assessment or remain unassessed following the 2012 status assessment due to data deficiencies or the unknown impact of hatchery produced fish within the CU (DFO 2016). 5 Phillips River is a new CWT indicator with consistent data time series starting with catch year 2014.

10 292 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Life Chinook History Escapement Management Unit Type CWT Indicators CTC Model Stock Indicators Associated Conservation Units4 CK-22: East Vancouver Island-Cowichan Lower Georgia & Koksilah Strait CK-25: East Vancouver Island-Nanaimo Ocean Cowichan (COW) Lower Georgia Strait Cowichan & Chemainus CK-31: West Vancouver Island-South CK-32: West Vancouver Island-Nootka & WCVI Natural Aggregate index West Coast Robertson Creek Kyuquot Ocean Vancouver Island (RBT) CK-33: West Vancouver Island-North WCVI Hatchery Aggregate index - CK-16: South Thompson-Bessette Aggregate Fraser run Fraser Spring 4 Stream Nicola (NIC) Fraser Spring 1.2 Creek_SU_1.2 2 reconstruction index CK-17: Lower Thompson_SP_1.2 CK-04: Lower Fraser River_SP_1.3 CK-06: Lower Fraser River_SU_1.3 Aggregate Fraser run CK-10: Middle Fraser River_SP_1.3 Fraser Spring 5 Stream - Fraser Spring 1.3 2 reconstruction index CK-12: Upper Fraser River_SP_1.3 CK-14: South Thompson_SU_1.3 CK-18: North Thompson_SP_1.3

CK-09: Middle Fraser River-Portage Aggregate Fraser run Fraser Summer 5 Stream - Fraser Summer 1.3 CK-11: Middle Fraser River_SU_1.3 2 reconstruction index CK-19: North Thompson_SU_1.3

Lower Shuswap (SHU) Aggregate Fraser run CK-07: Maria Slough_SU_0.3 Fraser Summer 4 Ocean Fraser Summer 0.3 CK-13: South Thompson_SU_0.3 1 Middle Shuswap reconstruction index

(MSH) CK-15: South Thompson_SU_0.3 Harrison River Fraser Fall 4 Ocean Harrison Fall Harrison River CK-03: Lower Fraser River_FA_0.3 1 (HAR) Chilliwack - Ocean Chilliwack Fall - CK-9008: Fraser-Harrison fall Hatchery (CHI) transplant_FA_0.3

11 293 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Table 2. Summary of recent trends in characteristics for select stocks within Transboundary and BC management units. WSP 2017 CU Escapement Integrated Generation Female 6 Survival Fecundity Escapement Trend Status Time Length Management Unit Stock Assessment (2007-2011 (% change over 3 (relative to Data to 2012 brood year avg generations up to (Decline rate) (Trend) (Trend) 2003-13 avg) (DFO 2016) relative to 1980- 2016) 1990 avg) Transboundary Alsek -65% - - Unk Unk Unk Unk Taku -73% - - -39% stable7 Declining Unk Stikine -71% - - Unk -0.026 Unk Unk Northern BC Nass -72% - - Unk Unk Declining Unk Skeena -68% - - -36% -0.025 Declining Unk Kitsumkalum -66% age-5,-6 Central BC Atnarko Total -13% - - 28% -0.015 Unk Unk Atnarko Wild -21% Upper Georgia NEVI NA NA CK-288 DD (CK-28) -81% -0.017 Declining Declining Strait (Quinsam) -60% CK-29 Red (CK-29) age-4,-5 since 2011 Big Qualicum -51% -45% CK-27 TBD -44% -0.017 Declining Declining age-3,-4 since 2011 Puntledge -45% -80% CK-83 TBD -9% -0.009 Unk Unk Summers Lower Georgia Cowichan 422% 386% CK-22 TBD -73% -0.008 Unk Stable Strait Nanaimo -5% CK-25 TBD Unk Unk Unk Unk WCVI WCVI aggregate 164% -12% CK-31 Red (CK-31) -73% stable Unk Unk 287% CK-32 Red (CK-32) -10% CK-33 TBD (CK-33)

Fraser Spring 42 Fraser Spring -52% -51% CK-16 Red (CK-16) -55% stable Declining, Unk 1.2 (Nicola) -67% 98% CK-17 Red (CK-17) age-4

6 Based on the short-term trend metric present in DFO (2016) and updated to include escapement time series ending in 2016 for all sites within the CU (wild and enhanced combined). Trends were estimated from a linear trend in loge(spawner abundances) over 3 generations, based on all years with reviewed escapement time series data and using infilled values where applicable. 7 The complete time series is stable, but shows a consistently declining trend since 1990. 8 CK-28 is represented by the Phillips River which is a relatively new CWT indicator (established in 2014) and as such, does not presently have a three generation time series to calculate trend.

12 294 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Fraser Spring 52 Fraser Spring -59% -44% CK-04 TBD (CK-04) Unk Unk Unk Unk 1.3 -43% CK-06 DD (CK-06) -3% CK-10 Red (CK-10) -53% CK-12 Red (CK-12) -45% CK-14 Red/Amber -79% CK-18 (CK-14) Red (CK-18) Fraser Summer Fraser Summer -68% -86% CK-09 Red (CK-09) Unk Unk Declining, Unk 52 1.3 (Chilko) -48% CK-11 Amber (CK- age-5 -72% CK-19 11) Red (CK-19) Fraser Summer Fraser Summer -21% -27% CK-07 TBD (CK-07) -42% -0.020 Declining Declining 41 0.3 -12% CK-13 Green (CK- age-3,-4,-5 (L Shuswap) -53% -36% CK-15 13) TBD (CK-15)

Fraser Fall 41 Fraser Fall -68% -28% CK-03 Green(p) -45% -0.016 Declining Unk (Harrison) 19% CK-9008 (CK-03) age-3,-4,-5 TBD (CK- 9008)

13 295 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Table 3. Average landed catch of Chinook in Southern BC marine fisheries, 1975 – 2016 (CTC 2017a). Note: Fraser River catch is not included in this table. Catch SBC ISBM SBC SBC SBC WCVI -WCVI Total % Change Years First ISBM ISBM ISBM AABM AABM Average Relative to Nations Net Sport Troll Troll Sport Catch 1975-84 marine9 1975 to 84 1,020 73,170 347,356 214,393 511,790 No Est 1,147,729 - 1985 to 98 7,201 34,710 167,438 26,300 221,065 20,572 477,286 -58% 1999 to 08 15,707 10,713 93,711 425 102,317 37,729 260,601 -77% 2009 to 16 12,189 11,229 97,324 0 70,080 58,494 249,315 -78%

Table 4. Average landed catch of Chinook in Northern and Central BC marine fisheries, 1975 – 2016 (CTC 2017a). Note: Transboundary catches are not included in this table. Catch Years NBC NBC NBC ISBM NBC NBC Total % Change ISBM ISBM Net Sport9 ISBM AABM Average Relative to First Troll Sport10 Catch 1975-84 Nations and Troll 1975 to 84 14,564 60,489 9,795 101,221 167,571 353,640 - 1985 to 98 27,038 40,600 10,432 26,594 154,977 259,641 -27% 1999 to 08 23,310 18,336 9,265 256 144,532 195,699 -45% 2009 to 16 14,529 6,797 9,237 0 146,369 176,932 -47%

9 Includes Food, Social, Ceremonial (FSC) and Economic Opportunities (EO) fisheries. 10 NBC Sport catches begin in 1977.

14 296 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Table 5. Estimated calendar year annual exploitation rates (CYER) (2011-2016) for Transboundary and BC Chinook management units relative to historic levels (1980-1989). Although CTC model output may be available for some of these stocks, only CWT Indicator Stock data is included below. See Appendix Table A-1. Annual exploitation rate estimates represent total Adult Equivalency (AEQ)-adjusted mortality (i.e. includes estimated release mortality). All estimates exclude youngest age fish (i.e., age-2 for ocean- type stocks, and age-3 for stream-type stocks). 11

Average CYER Average CYER Management Unit CWT Indicator Stock 1980-89 2011-2016 % Change Transboundary Stikine - 23% - - Taku 11% 16% 43% Northern BC Kitsumkalum River 43% 33% -24% Central BC Atnarko River - 40% - Phillips - 26% - Quinsam Hatchery 71% 37% -48% Upper Georgia Strait Puntledge Hatchery 63% 42% -32% Big Qualicum Hatchery 73% 44% -39% Lower Georgia Strait Cowichan River - 62% - Robertson Creek WCVI 56% 35% -37% Hatchery

Fraser Spring 42 Nicola River - 20% -

Fraser Spring 52 None currently - - -

Fraser Summer 52 None currently - - - Lower Shuswap River 35%12 44% 26% Fraser Summer 41 Middle Shuswap - 44% - Harrison River 75%13 28% -62% Fraser Fall 41 Chilliwack River 69% 28% -59%

11 Table 5 has been revised from an earlier version of this SR; see Appendix table A-1 for comparison and explanation. 12 Based on 1988-1989 only. 13 Based on 1985-1989 only.

15 297 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Table 6. Summary of the intrinsic productivity (Ricker α) and sustainable exploitation rates (EMSY) for select management units based on Dorner et al. (2017; data available from B. Dorner upon request). Note: Dorner et al (2017) excludes hatchery stocks. Chinook productivity analysis for their full time series and for the recent five brood years up to 2008 for B.C. stocks. Also displayed are recent CWT exploitation rates for Transboundary and B.C. Chinook management units, expressed in brood years (BYER) and catch (calendar) years (CYER) 14. Further work is required to estimate long term and recent productivity for several B.C. management units following the method applied by Dorner et al. (2017).

Long term Recent Relative Relative Recent Average productivity productivity change Long term Recent change BYER CYER Management Unit Indicator Stock (Ricker α) (Ricker α) in prod. EMSY EMSY EMSY 2004-2008 2011-2016 Transboundary Alsek 0.74 0.25 -66% 33% 12% -64% - - Stikine 1.45 1.50 3% 58% 59% 2% 34% 23% Taku 0.94 0.70 -26% 41% 31% -23% 24% 16% Northern BC Kitsumkalum 1.51 1.28 -15% 60% 53% -12% 32% 33% Central BC Atnarko ------51% 40% 15 Upper Georgia Strait Phillips ------26% Quinsam ------41% 37% Puntledge ------37% 42% Big Qualicum ------44% 44% Lower Georgia Strait Cowichan ------76% 62% WCVI Robertson ------37% 35%

Fraser Spring 42 Nicola ------27% 20%

Fraser Spring 52 None currently ------Fraser Summer 52 None currently ------

Fraser Summer 41 Lower Shuswap ------50% 44% - Middle Shuswap ------55%16 44%

Fraser Fall 41 Harrison 1.18 0.59 -50% 49% 27% -45% 30% 28%

14 Estimates of Brood Year Exploitation Rate (BYER) and Calendar Year Exploitation Rate (CYER) exclude the youngest age classes (i.e. age-2 for ocean type stocks, and age-3, for stream type stocks). BYER is based on complete brood years (with the same input years as EMSY) and thus is comparable to EMSY. CYER can be estimated for incomplete brood years and can provide an early indication of recent changes in exploitation if it differs from BYER (though it is not in the same “currency” as BYER or EMSY). 15 Philips CYER is only based on 2014-2016. 16 Middle Shuswap BYER is based on 2008 value only and CYER is based on 2014-2016.

16 298 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Table 7. Escapement for indicator stocks relative to CTC-accepted escapement goals derived from stock- recruit analyses (S-R) or habitat-based estimates of SMSY (estimated spawners required to produce maximum sustained yield, SMSY; Parken et al. 2006) for Transboundary and BC management units. Note: in all cases there is significant uncertainty in the estimated SMSY values and for most management units; the estimates are preliminary and require further review. Estimates of escapement include both hatchery and wild adult contributions, but exclude estimates of jacks.

Escapement Goal Escapement17 Management Unit Stock Long-term Last 5 Lower Upper Type 2017 AVG Years 8,586 Alsek 3,500 5,300 S 3,672 1,740 MSY (1976-2017) 36,140 Transboundary Taku 19,000 36,000 S 18,304 8,754 MSY (1975-2017) 24,635 Stikine 14,000 28,000 S 15,997 7,206 MSY (1976-2017) 17,344 Nass 10,000 15,000 S-R 11,411 4,984 (1977-2017) Northern BC 13,764 Kitsumkalum 8,621 - S 10,225 4,943 MSY (1984-2017) 17,591 Central BC Atnarko 5,009 - S 24,219 10,395 MSY (1990-2017)

Upper Georgia Strait Aggregate ------

5,491 18 Lower Georgia Strait Cowichan 6,500 - S-R 6,590 10,590 (1981-2017) 11,304 WCVI 15,000 - S 17,727 17,163 Aggregate MSY (1993-2017) 10,693 Fraser Spring 4 22,146 - S 11,317 5,105 2 Aggregate MSY (1975-2017) 23,805 Fraser Spring 5 Aggregate 42,165 - S 18,916 8,154 2 MSY (1975-2017) 20,047 Fraser Summer 5 23,567 - S 16,070 6,459 2 Aggregate MSY (1975-2017) 63,006 Fraser Summer 4 120,000 322,000 S 111,950 84,470 1 Aggregate MSY (1975-2017) 94,958 Fraser Fall 4 Harrison 75,100 98,500 S-R 52,056 29,799 1 (1984-2017)

17 Although jacks can provide an indication of future abundance in situations where a sibling relationship exists, decreasing maturation trends have led to changes in the sibling relationships for some stocks (i.e., older ages consistently returning below that expected from the sibling regression relationship). Preliminary observations show that there are regional differences in the recent patterns of jack returns (i.e., Vancouver Island stocks generally show an increasing trend while Fraser River stocks show a decreasing trend). 18 Excludes broodstock removals.

17 299 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Figure 1. Trends in combined hatchery and natural escapement based on average deviations (Z-scores) for 12 Chinook management units, 1975-2017. Stocks included in the analysis are Transboundary: Alsek, Taku, Stikine; North Coast: Nass, Skeena, Kitsumkalum; Central Coast: Atnarko Total; NWVI: Artlish, Tahsis, Kaouk, Tahsish; SWVI: Megin, Moyeha, Bedwell; Upper Georgia Strait: Nimpkish, Salmon, Quinsam/Campbell; Lower Georgia Strait: Cowichan; Fraser aggregates for Spring 42, Spring 52, Summer 52 and Summer 41 and Fraser Fall 41: Harrison.

18 300 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Figure 2a. Trends in brood year marine survival (commonly represented by cohort-based smolt to age-2 or -3 survival, CTC 2017b) estimated for nine CWT indicator stocks across Transboundary and BC rivers (excluding the Fraser River). Note: Estimates for incomplete broods are depicted by grey bars; these estimates are calculated using average maturation rate assumptions and are expected to change until fish have matured at all possible ages. 2015 brood year observations (which use model estimates for all but the age-2 class) have been excluded. Estimates include CWT recoveries from 2017. Data for each stock have been standardized by natural log transformation and scaled to a mean of zero and standard deviation of one.

19 301 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Figure 2b. Trends in brood year marine survival (commonly represented by cohort smolt to age-2 or -3 survival, CTC 2017b) estimated for four CWT indicator stocks from the Fraser River. Note: Estimates for incomplete broods are depicted by grey bars; these estimates are calculated using average maturation rate assumptions and are expected to change until fish have matured at all possible ages. 2015 brood year observations (which use model estimates for all but the age-2 class) have been excluded. Estimates include CWT recoveries from 2017. Data for each stock have been standardized by natural log transformation and scaled to a mean of zero and standard deviation of one.

20 302 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Figure 3a. Observed changes in average generation time for Transboundary and (non-Fraser) BC CWT indicator stocks in the absence of fishing. Calculated values are based on CWT recoveries up to and including 2016. Slope value estimates the rate of change in average generation time over the time series. P-val provides a statistical measure of the probability the slope is significantly different from zero.

21 303 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Figure 3b. Observed changes in average generation time for Fraser River CWT indicator stocks, in the absence of fishing, brood years 1980-2011. Calculated values are based on CWT recoveries up to and including 2016. Slope value estimates the rate of change in average generation time over the time series. P-val provides a statistical measure of the probability the slope is significantly different from zero.

22 304 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Figure 4a. Trends in total annual simple calendar year exploitation rate for Transboundary and (non- Fraser) BC Chinook CWT indicator stocks. The exploitation rate represents total mortality (i.e., includes estimated incidental mortality from releases as well as landed catch). The estimates exclude youngest age fish and have been adjusted for age-specific adult equivalency. For the Robertson Creek Hatchery stock, estimates are based on pre-terminal fisheries only as terminal fisheries targeting the hatchery returns in Alberni Inlet are not representative of terminal impacts on natural stocks in the WCVI region.

23 305 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Figure 4b. Trends in total annual simple calendar year exploitation rate for Fraser River Chinook CWT indicator stocks. The exploitation rate represents total mortality (i.e., includes estimated incidental mortality from releases as well as landed catch). The estimates exclude youngest age fish and have been adjusted for age-specific adult equivalency.

24 306 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

Appendix Table A-1. Unrevised (original) Table 5: estimates of calendar year annual exploitation rates (2011- 2016) for Transboundary and BC Chinook management units relative to historic levels (1980-1989) provided in previous SR version. In the original (unrevised) Table 5, estimates from modelled outputs from the Chinook Technical Committee coast-wide model were provided for Transboundary, Fraser Spring 52 and Fraser Summer 52.stocks. The unrevised Table 5 estimates included youngest age fish (i.e., age-2 for ocean-type stocks, and age-3 for stream-type stocks), and escapement strays. Table A-1 is included as a comparison only to illustrate the revisions in the updated Table 5 above.

CWT Indicator Average ER Average ER Management Unit % Change Stock / Model Stock 1980-89 2011-2016

Transboundary CTC Model Output 24% 39% 66%

Northern BC Kitsumkalum River 46% 35% -23%

Central BC Atnarko River n/a 40% n/a

Cowichan River 79% 55% -30%

Big Qualicum 76% 43% -43% Hatchery Inner South Coast Puntledge Hatchery 66% 37% -43%

Quinsam Hatchery 73% 38% -48%

Robertson Creek WCVI 56% 35% -37% Hatchery

Fraser Spring 42 Nicola River 33% 18% -44%

Fraser Spring 52 CTC Model Output 36% 26% -28%

Fraser Summer 52 CTC Model Output 51% 35% -31% Lower Shuswap 37% 42% 16% River Fraser Summer 41 Mid Shuswap n/a 46% n/a

Fraser Fall 41 Harrison River 73% 24% -67%

25 307 Science Response: Science Information to support Pacific Region Chinook Salmon management measures 2018

This Report is Available from the : Centre for Science Advice Pacific Region Fisheries and Oceans Canada 3190 Hammond Bay Road Nanaimo, BC V9T 6N7 Telephone: (250) 756-7208 E-Mail: [email protected] Internet address: www.dfo-mpo.gc.ca/csas-sccs/ ISSN 1919-3769 © Her Majesty the Queen in Right of Canada, 2018

Correct Citation for this Publication: DFO. 2018 Science information to support consultations on BC Chinook Salmon fishery management measures in 2018. DFO Can. Sci. Advis. Sec. Sci. Resp. 2018/035. Aussi disponible en français : MPO. 2018. Information scientifique à l’appui des consultations sur les mesures de gestion des pêches au saumon quinnat de la Colombie-Britannique (2018). Secr. can. de consult. sci. du MPO, Rép. des Sci. 2018/035.

26 308 Oil-Particle Interactions and Submergence from Crude Oil Spills in Marine and Freshwater Environments— Review of the Science and Future Science Needs

Open-File Report 2015–1076

U.S. Department of the Interior U.S. Geological Survey 309 310 Oil-Particle Interactions and Submergence from Crude Oil Spills in Marine and Freshwater Environments—Review of the Science and Future Science Needs

By Faith A. Fitzpatrick, Michel C. Boufadel, Rex Johnson, Kenneth Lee, Thomas P. Graan, Adriana C. Bejarano, Zhenduo Zhu, David Waterman, Daniel M. Capone, Earl Hayter, Stephen K. Hamilton, Timothy Dekker, Marcelo H. Garcia, and Jacob S. Hassan

Open-File Report 2015–1076

U.S. Department of the Interior U.S. Geological Survey

311 U.S. Department of the Interior SALLY JEWELL, Secretary U.S. Geological Survey Suzette M. Kimball, Acting Director

U.S. Geological Survey, Reston, Virginia: 2015

For more information on the USGS—the Federal source for science about the Earth, its natural and living resources, natural hazards, and the environment—visit http://www.usgs.gov or call 1–888–ASK–USGS. For an overview of USGS information products, including maps, imagery, and publications, visit http://www.usgs.gov/pubprod/.

Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. Although this information product, for the most part, is in the public domain, it also may contain copyrighted materials as noted in the text. Permission to reproduce copyrighted items must be secured from the copyright owner.

Suggested citation: Fitzpatrick, F.A., Boufadel, M.C., Johnson, Rex, Lee, Kenneth, Graan, T.P., and others, 2015, Oil-particle interactions and submergence from crude oil spills in marine and freshwater environments—Review of the science and future science needs: U.S. Geological Survey Open-File Report 2015–1076, 33 p., http://dx.doi.org/10.3133/ofr20151076.

ISSN 2331-1258 (online)

312 iii

Acknowledgments

A synthesis paper like this, even one with a long author list, doesn’t do justice to the amount of effort and time spent on each of the studies mentioned in this report. Special thanks are given to U.S. Environmental Protection Agency and Enbridge Energy, L.P., contractors and collaborators involved in assessment and monitoring as well as recovery and containment at the Kalamazoo River site.

313 Contents Abstract ...... 1 Introduction ...... 1 Purpose and Scope ...... 4 Review of the Science ...... 4 Formation of Oil-Particle Aggregates ...... 4 Oil-Particle Aggregates as a Natural Physical Dispersant ...... 11 Transport and Fate of Oil-Particle Aggregates ...... 13 Ecological Risk and Toxicity of OPAs and Oiled Sediment ...... 15 Effects of Ice in Northern Climates on OPA Formation and Spill Response ...... 17 Operational Considerations ...... 18 Detection ...... 18 Containment ...... 20 Recovery ...... 21 Future Science Needs ...... 24 Summary and Conclusions ...... 24 References Cited ...... 25

iii

314 Figures 1. Simplified diagram of the processes and mechanisms leading to oil-particle aggregate formation and breakup in marine and freshwater environments...... 2

2. Summary of American Petroleum Institute (API) gravities and dynamic viscosities (at approximately 10-20 degrees Celsius) for major categories of crude oil types and bitumen/oil sands...... 6 3.Types of oil-particle aggregates: A, single and multiple droplet aggregate, B, solid aggregate of large, usually elongated mass of oil with interior particles (dashed blue circles), and C, flake aggregate of thin membranes of clay aggregates that incorporate oil and fold up (modified from Stoffyn-Egli and Lee, 2002)...... 8 4. Salinity in relation to the percent of oil in oil-particle aggregates (OPAs) from laboratory shaker tests of diluted heavy crude oil form the 2000 OSSA II spill into the Rio Desaguadero in Bolivia and its smectite-rich sediment...... 8

5. Streamflow in relation to suspended sediment concentration in the Kalamazoo River, Michigan, August 2012-March 2014………...... 10

6. Cumulative particle-size distribution for suspended sediment collected April 22, 2013, during flows of 32 cubic meters per second (m3/s) at Marshall, Michigan, and 85 m3/s near Battle Creek, Michigan ...... 10

7. Kalamazoo River sediment spiked with weathered source oil after 48 hours, under ultraviolet-epifluorescence microscopy at 320 times magnification...... 11

8. Flowchart used for field observations of submerged oil during poling assessments ...... 21

9. A, Diagram of subsurface curtain boom, and B, map of containment boom locations for submerged oil in Morrow Lake Delta from river miles 37.25 to 37.75, Kalamazoo River, Michigan related to the Enbridge Line 6B oil release……...... 22 Tables

1. Nonparametric Spearman's Rho correlation coefficients between acute sediment toxicity test results and analytical variables using Chironomus dilutus and Hyalella azteca...... 16

315 Conversion Factors SI to Inch/Pound Multiply By To obtain Length micrometer (m) 0.000039 inch (in.) centimeter (cm) 0.3937 inch (in.) millimeter (mm) 0.03937 inch (in.) meter (m) 3.281 foot (ft) kilometer (km) 0.6214 mile (mi) Area square meter (m2) 0.0002471 acre square kilometer (km2) 247.1 acre square centimeter (cm2) 0.001076 square foot (ft2) square meter (m2) 10.76 square foot (ft2) square kilometer (km2) 0.3861 square mile (mi2) Volume liters (l) 0.0063 Barrel (petroleum, 1 barrel = 42 gal) liters (l) 0.2642 gallon (gal) cubic meter (m3) 6.290 barrel (petroleum, 1 barrel = 42 gal) cubic meter (m3) 35.31 cubic foot (ft3) cubic meter (m3) 1.308 cubic yard (yd3) Flow rate meter per second (m/s) 3.281 foot per second (ft/s) cubic meter per second (m3/s) 35.31 cubic feet per second (ft3/s) Density

gram per cubic centimeter (g/cm3) 62.4220 pound per cubic foot (lb/ft3)

Stress

pascal (Pa) 0.000145 pound per square inch (psi)

Dynamic viscosity centipoise (cP) 2.42 pound/foot-hour (lb/ft-hr) Energy dissipation rate square meters per cubic second (m2/s3) 10.76 square feet per cubic second (ft2/s3)

Temperature in degrees Celsius (°C) may be converted to degrees Fahrenheit (°F) as follows: °F=(1.8×°C)+32

316 Oil-Particle Interactions and Submergence from Crude Oil Spills in Marine and Freshwater Environments—Review of the Science and Future Science Needs

By Faith A. Fitzpatrick1, Michel C. Boufadel2, Rex Johnson3, Kenneth Lee4, Thomas P. Graan5, Adriana C. Bejarano6, Zhenduo Zhu7, David Waterman7, Daniel M. Capone8, Earl Hayter9, Stephen K. Hamilton10, Timothy Dekker11, Marcelo H. Garcia7, Jacob S. Hassan12

Abstract Oil-particle interactions and oil submergence are of much interest to oil spill responders and scientists, especially as transportation of light and heavy crude oils increases in North America’s coastal marine and freshwater environments. This report contains an up-to-date review of the state of the science for oil-particle aggregates (OPAs), in terms of their formation and stability which may alter the transport, fate, and toxicity of the residual oil and, hence, its level of ecological risk. Operational considerations—detection, containment, and recovery—are discussed. Although much is known about oil-particle interactions in coastal marine environments, there remains a need for additional science on methods to detect and quantify the presence of OPAs and to understand their effects on containment and recovery of oil spilled under various temperature regimes and in different aquatic habitats including freshwater environments. Introduction Suspended particles affect the fate and transport of spilled oil in aquatic environments (Muschenheim and Lee, 2002; Owens and Lee, 2003; Khelifa and others, 2005a, b, c; Sun and Zheng, 2009; Gong and others, 2013), such as rivers and floodplains, shorelines and beaches along lakes and oceans, coastal and riparian wetlands, and deeper waters of oceans and lakes. The manner in which oil interacts with particles and its eventual transport and fate depend on the physical properties of the oil and the particles, as well as environmental conditions including the geomorphic setting, weather, currents, and vertical mixing of the water column (Lee, and others, 2011a; Lee and others, 2002) (fig. 1). Combinations of oil and particles have various names, including clay-oil flocculation (Bragg and Yang, 1995), oil-mineral aggregates (Lee and others, 1998) and oil-suspended sediment-aggregates (Khelifa and others, 2002) depending on the type of particle involved in the interaction. The term oil- particle aggregate (OPA) is used in this report because it is the more generic term that includes a wider range of particles containing both mineral sediment and organic matter in association with oil that may be retained in suspension and (or) settled out.

1U.S. Geological Survey Wisconsin Water Science Center, 2New Jersey Institute of Technology, 3Global Remediation Technologies, Inc., 4Commonwealth Scientific Industrial Research Organization (CSIRO), 5Weston Solutions, Inc., 6Research Planning, Inc., 7University of Illinois Ven Te Chow Hydrosystems Laboratory, 8Mannik Smith Group, 9U.S. Army Corps of Engineers, Engineer Research and Development Center, 10Michigan State University, 11LimnoTech, Inc. 12U.S. Environmental Protection Agency

317

Figure 1. Simplified diagram of the processes and mechanisms leading to oil-particle aggregate formation and breakup in marine and freshwater environments. (modified from Environment Canada, 2013; OPA, oil-particle aggregate).

Traditional clean-up methods based on physical recovery of oil slicks on surface waters, such as booming and skimming, are ineffective for spilled oil once it submerges. Observations of OPA formation associated with the transfer of oil to the benthic environment (Lee, 2002; Payne and others, 2003; Sterling and others, 2005; Passow and others, 2012) have provided the justification for additional scientific studies to understand the processes and characteristics associated with OPA formation and their transport and fate. Results from additional studies will likely influence oil spill response contingency planning and spill response operations that include methodologies for detection, monitoring, recovery, and containment (Bandara and others, 2011; Environment Canada, 2013; Dollhopf and others, 2014; Hansen, 2014). OPA formation has been correlated with the removal of oil stranded within sediments in freshwater and marine environments by both natural recovery (natural attenuation) and active remediation procedures (the addition of suspended particulate material in the presence of turbulent mixing energy) (Bragg and Yang, 1995; Lee and others, 1996; 1997; Wood and others, 1997; Lee, 2002; Lee and others, 2002; Owens and Lee, 2003). Lee and others (1997; 2003a) note that OPA formation accelerated the removal of stranded oil within the intertidal/surf zone by reducing the adhesive properties of the oil and the tendency of dispersed droplets stabilized by mineral fines to re- coalesce. Thus, either formed naturally or enhanced with addition of clay minerals, the relatively stable OPAs are dispersed more easily in the water column, potentially reducing the oil to concentrations below toxicity threshold limits (Lee and others, 2003a; Lee and others, 2003b) and making the oil more available for biodegradation (Weise and others, 1999; Lee and others, 1996; Lee and others, 1997; Lee and Merlin, 1999; Owen and Lee, 2003). Like chemical dispersants, the exposure pathway is altered

318 from water surfaces and shorelines to the water column, which transfers the toxicity risks from water fowl and shoreline organisms to planktonic, open water, and benthic species (Venosa and others, 2014). Submergence can happen to light and heavy oils after they bind to relatively small amounts of particles (mineral sediment or organic matter). This can occur in suspension or while the oil moves along the bottom of a water body, bank, or coastline (Lee and others, 1985; Lee, 2002; Cloutier and others, 2003). The behavior of bitumen from the Canadian tar sands (usually more dense and viscous than traditional heavy oil), and particularly its interaction with particles leading to eventual submergence and accumulation of OPAs in sediments, has recently become a topic of much interest because of the 2010 Enbridge Line 6B pipeline release into the Kalamazoo River when more than 3.2 million liters of diluted bitumen were released into the Kalamazoo River in Michigan (Dollhopf and others, 2011). Response operations for that release began with conventional recovery techniques for floating oil and recovered about 2.9 million liters; however, submerged oil became the focus approximately 1 month into the cleanup and remained the focus through the summer of 2014 (Enbridge Energy, L.P., 2010; Dollhopf and others, 2014). Given the timely research efforts pertaining to the Canada Northern Gateway proposed pipeline, recent studies provided insights on the buoyancy (King and others, 2014), interaction with mineral fines (King and others, 2015), and the ultimate fate of OPAs formed from bitumen (Environment Canada, 2013). Light crude oil can interact with particles along rivers, as indicated by the 2013 derailment, explosion, and spill of light crude oil in Lac-Mégantic, Quebec when an estimated 100,000 liters of oil spilled into the Chaudière River. A management plan by the Government of Quebec was developed that included recommendations for cleanup of river bottom sediment contaminated with hydrocarbons (Gouvernement du Quebec, Depot Legal, 2014). Laboratory tests at Louisiana State University using E2MS 303 oil from the February 2014 spill of Bakken crude from a barge collision into the lower Mississippi River indicate that the “oil will quickly adhere to suspended solids in the water column, forming unstable emulsions” (Doelling and others, 2014). More than 15 years ago, before the emergence of concerns about increased pipeline transport of diluted bitumen, the National Coastal Research Council (on behalf of the U.S. Coast Guard) published “Spills of Nonfloating Oils: Risk and Response,” a report that included specific recommendations for detection, monitoring, modeling, and recovery of submerged oil, mainly in marine environments (National Research Council, 1999). These recommendations were further tabulated into science needs for detection and monitoring, fate and transport, containment and recovery, and effects and restoration and included specific mention of OPAs (Coastal Response Research Center, 2007). Some areas of science support included developing better sensors to detect OPAs, mapping the extent of OPAs, and determining future resuspension and remobilization of OPAs in bottom sediment. The Coastal Response Research Center (2007) noted that improved characterization of the size, composition, and distribution of particles would enable better forecasting, observation, understanding, and hind casting of OPA behavior for a range of ecological and geomorphic settings. Fisheries and Oceans Canada has been conducting oil-particle interactions studies with oil and chemically dispersed oils since the mid-1980s. An oil-particle interactions workshop was organized in 2000, and a compendium of papers appeared in the Spill Science and Technology Bulletin in 2003 and 2004 (Lee and Jarvis, 2004). These publications formed the basis for a quickly growing body of literature on the topic of OPAs in both marine and freshwater environments. Because oil transportation in North America increased greatly on rails, through pipelines, and on barges and vessels as a result of the increased production of Bakken crude in the Williston Basin, North Dakota, and bitumen in the western provinces of Canada (Frittelli and others, 2014; Association of American Railroads, 2014; Committee for a Study of Pipeline Transportation of Diluted Bitumen,

319 2013), more attention is being given to submerged oil detection and response techniques in marine and freshwater environments (Coastal Response Research Center, 2007; Hansen, 2014). It is clear that more science is needed on the formation, settling, resuspension, and toxicity of OPAs, and the manner in which these properties vary depending on the nature of the oil and the characteristics of the aquatic environment it enters. Quantification of residual oil following spill response operations needs to account for the fraction of oil associated with suspended/settled particulate material.

Purpose and Scope The purpose of this report is to describe the state of the science concerning OPAs—their formation, transport, settling, resuspension, and breakup in a variety of geomorphic environments including coastal marine, deep ocean, freshwater lacustrine (Great Lakes), and riverine, and brackish river mouths and harbors. The report contains a synthesis and review of available literature on OPAs from laboratory, experimental tank and flume studies, and some large-scale field experiments. Ongoing studies are described and needs for continuing investigation and new science on OPAs are listed. Review of the Science The U.S. Geological Survey (USGS) mission is to provide reliable scientific information to describe and understand the Earth; minimize loss of life and property from natural disasters; manage water, biological, energy, and mineral resources; and enhance and protect our quality of life. That mission has involved the USGS in providing science support for oil spill recovery. This review of the science on oil-particle interactions was motivated by questions that have arisen during oil spill response of the Enbridge Line 6B spill in Marshall, Michigan. However, the review is holistic in the sense that it incorporates spills of oils with various properties in freshwater and marine environments, as a result of concerns about spills from the rising quantities of crude oil produced and transported across the Nation. The following questions helped to form the major sections included in this review:  What environmental conditions lead to the formation of OPAs?  What is the effectiveness of adding particles to an oil spill for physical dispersion of oil as a spill countermeasure?  What is the long-term fate and transport of OPAs?  What are the ecological implications (fate and effects) of OPAs?  Are there special circumstances for OPAs in cold climates for use as a spill countermeasure or for recovery in icy water?  What are the operational considerations for recovery of OPAs?

Formation of Oil-Particle Aggregates Formation of OPAs happens naturally when oil and suspended particles mix in turbulent water (Lee, 2002; Muschenheim and Lee, 2002; Owens and Lee, 2003; Khelifa and others, 2005a; Sun and Zheng, 2009; Gong and others, 2013). An understanding of the exact mechanisms of OPA formation comes mainly from the use of laboratory shaker and wave tanks, including those at the Bedford Institute of Oceanography, Dartmouth, Nova Scotia (Center of Offshore Oil and Gas Environmental Research, 2004), and flume experiments. Results from wave tank and laboratory shaker experiments provide empirical data for models of OPA formation and breakup (Stoffyn-Egli and Lee, 2002; Li and others, 2007; Ma and others, 2008; Wang and others, 2011). However, field observations and data from

320 assessment and monitoring activities at spill sites yield information on the timing and range of environmental conditions under which OPAs have formed (Niu and others, 2011). Major factors affecting the formation of OPA are (1) quantity and viscosity of the oil, interfacial tension of oil-water, and chemical composition of the oil; (2) quantity, type, and surface properties of the particles; (3) magnitude and variability in physical energy of the aquatic environment; (4) temperature; and (5) salinity (Lee, 2002; Khelifa and others, 2002; Payne and others, 2003). Recent tests have focused on the highly viscous diluted bitumen (dilbit) product (Zhao and others, 2014a). In spills of heavy crude oil, particularly bitumen, freshwater environments with fine-grained sediment in the water column and bottom, abundant sunlight, warm temperatures, and strong currents and turbulence create a high potential for oil submerging and ultimately being deposited in the sediments (Silliman, 2014). The first step to forming OPAs lies with the initial breakup of a slick of oil into oil droplets. Once spilled into a water body with turbulence created by waves or currents, floating oil can break up into droplets and reach a stable droplet size distribution (DSD) relatively quickly, perhaps in minutes to tens of minutes (Zhao and others, 2014b). Smaller droplets are generated when the interfacial tension of oil-water is small and (or) the oil viscosity is small. The interfacial tension of oil-water is more or less constant in the absence of surfactants. However, oil viscosity can increase by orders of magnitude among different types and temperatures of oils (fig. 2). For example, the viscosity of a heavy crude oil or bitumen is at least 1,000 times that of light crude such as a product from the Alaska North Slope or Bakken Formation oil (http://www.etc-cte.ec.gc.ca/databases/oilproperties/Default.aspx). Figure 2 contains a summary of American Petroleum Institute (API) gravity and viscosity for the major categories of crude oil (American Petroleum Institute, 2011). Crude oil densities are usually measured in terms of degrees API gravity. A higher API gravity corresponds with lighter density, with freshwater at 10 degrees API. Crude oils are generally further categorized by their type of hydrocarbon base (paraffinic or naphthenic) and sulfur content (high–sour, low–sweet). For the common numerical models for the evolution of the DSD, it has been assumed that the interfacial tension is the only force resisting the breakup of droplets (Prince and Blanch, 1990; Tsouris and Tavlarides, 1994), which is not the case for high viscosity oils or for situations when surfactants (in the form of chemical dispersants) are used (Wang and Calabrese, 1986). Alternatively, Delvigne and Sweeney (1988) developed an expression that predicts dispersion, which is based on oil viscosity without any information on the oil-water interfacial tension. Thus, the Delvigne and Sweeney (1988) formula cannot be relied upon to account for the reduction in the interfacial tension when dispersants are used. Recently, Zhao and others (2014b) developed a comprehensive model that accounts for resistance to breakup from both interfacial tension and viscosity of the oil. This model is called VDROP to stress the important role of viscosity. The DSD depends also on the mixing energy, namely the dissipation rate of kinetic energy (National Research Council, 2005; Kaku and others, 2006). Thus, high mixing energy promotes the breakup of droplets (fig. 1). In addition, dilution (as oil mixes with a larger volume of water away from the source) increases the distance between oil droplets and thus minimizes the probability of collision and subsequently the coalescence of oil droplets into larger droplets.

321 50

API Bakken crude 40

Syncrude Light crude 30 Husky Synthetic Blend Medium Western Canadian Select degrees crude in

20 Heavy crude Cold Lake

gravity, 10 Extra heavy crude Bitumen API 0 1 10 100 1,000 10,000 100,000 Dynamic viscosity, in centipoise

Figure 2. Summary of American Petroleum Institute (API) gravities and dynamic viscosities (at approximately 10- 20 degrees Celsius) for major categories of crude oil types and bitumen/oil sands. Overlays are of example crude oils commonly transported in North America. Data summarized from Attanasi and Meyer (2007), American Petroleum Institute (2011), Sia Partners Energy Outlook (2011), Environment Canada’s Oil Properties Web Site http://www.etc-cte.ec.gc.ca/databases/oilproperties/Default.aspx, Andrews (2014), Doelling and others (2014), Crude Quality, Inc. (2014); and Enbridge Energy Partners, L.P. (2013).

The size of the oil droplet is a very important factor in the transport and fate of the oil and its interaction with particles. Because large droplets have higher buoyancy than smaller droplets, they tend to float to the water surface, whereas smaller droplets could be driven more easily in the water column as a result of mixing energy (Boufadel and others, 2007). Smaller droplets also have a larger specific area than large droplets, which would enhance dissolution and microbial biodegradation (Lee and others, 1997; Lee and Merlin, 1999; Reddy and others, 2012; Geng and others, 2013). When particles are attached to oil droplets, they form a type of barrier around the droplet that keeps it from coalescing with other oil droplets (Khelifa and others, 2005a). Droplets with particles result in three main features that contribute to the eventual fate and submergence of droplets: (1) OPA- enhanced stability of the oil droplets that limits coalescence; (2) increased specific density of coated droplets which causes dispersion into a water body and possible settling; and, (3) enhanced microbial degradation. The actual attachment of oil droplets to particles depends on the viscosity and adhesion properties of the oil droplet and the surface area of the particles, as well as salinity of the water body (Lee, 2002; Khelifa and others, 2005a). There is evidence that salinity increases the formation of OPAs as a result of the reduction of the thickness of the double layer with increased ionic strength of water (Clark, 2009). However, OPA was also observed to form in freshwater (Lee, 2002; Lee and others, 2002). If the droplet size distribution of oil has not reached a steady state prior to interaction with the

322 particles, the interaction would affect the breakage mechanism of oil droplets. The reduction in the net interfacial tension resulting from attachment of particles would make the oil viscosity the main force resisting oil breakup. In such a case, a model that accounts for oil viscosity in resisting breakup is needed, such as VDROP (Zhao and others, 2014b). Particle size, amount, and type are important to the formation of OPAs. Clay-sized mineral particles are effective at forming OPAs, especially if the particles are present in large concentrations (Lee and others, 1998; Lee, 2002). However, most natural waters have a range of particle sizes or at least bimodal distributions, and natural particle-size distributions also may vary with the amount of mixing resulting from waves or currents (Boufadel and others, 2007). The particles may be smaller than the oil droplet and form a coating, or they may be larger or the same size for an aggregate. In either droplet type (fig. 3A), the attachment of the oil droplet and particle reduces the interfacial tension of the OPA with the water, and thus the particles have an overall effect similar to surfactants (Lee, 2002). Oil also may attach to mineral or organic matter (Lee, 2002). Phytoplankton can readily form oil-organic aggregates in laboratory tests (Lee and others, 1985). Particle concentrations can be relatively low and still form OPAs (Lee and Stoffyn-Egli, 2001; Rymell, 2009); Khelifa and others (2002) note that, in laboratory shaker experiments with seawater, mineral concentrations as low as 100 milligrams per liter (mg/L) and moderate shaking resulted in OPA formation with various types of oil. The shape and makeup of OPAs can take multiple forms. Methods used to study the shape and makeup of OPAs, which are generally less than 1 millimeter (mm), involve instrumentation such as combined transmitted light/ultraviolet (UV) light epi-fluorescence microscopy, confocal scanning laser microscopy, and environmental scanning electron microscopy (Stoffy-Egli and Lee, 2002; Lee and others, 2012). The most common forms of OPAs involve a spherical oil droplet surrounded by particles or multiple spherical droplets in a particle aggregate (Bragg and Owens, 1994; Stoffyn-Egli and Lee, 2002; Khelifa and others, 2005a; Zhang and others, 2010; Lee and others, 2012) (fig. 3A). Spherical shaped OPA indicate that the spilled oil formed droplets before forming aggregates, as described in the previous paragraphs. Stoffyn-Egli and Lee (2002) also found two other types of aggregates in laboratory experiments—solid and flake types (figs. 3B and 3C)—that do not necessarily form from dispersed droplets. These additional types are distinguished in that the oil takes on more of the form of the particle or mineral, and the flake type is distinguished by having folds or rolls in a dendritic or feather shape. For all types, the combination of oil and particles can result in a range of specific gravities that are usually similar to, or heavier than, water; thus OPAs can be floating, neutrally buoyant (in suspension), or negatively buoyant (submerging or settling) (Stoffyn-Egli and Lee, 2002). Sometimes the mineral makeup of particles determines their buoyancy, and Omotoso and others (2002) found that low-viscosity oils formed negatively buoyant OPAs with hydrophilic minerals (having a strong affinity for water) but formed positively buoyant OPAs with calcite minerals. Lastly oil-particle interactions can be enhanced by colloidal mechanisms of coagulation of ions (Lee and Stoffy-Egli, 2001), as well as biological activity associated with bacteria and phytoplankton (Passow and others, 2012). Laboratory simulations of the formation of OPAs in a brackish (salinity of 1.5 parts per thousand) high-energy riverine environment (Rio Desaguadero) were done by mixing sediment-laden water and heavy crude oil from the 2000 OSSA II pipeline spill into the Rio Desaguadero in the Bolivian Altiplano (Lee and others, 2001; 2002). In this spill, a missing oil fraction of 27–37 percent was not recovered; presumably it succumbed to from oil dispersion and enhanced biodegradation caused by the formation of OPAs. The river sediment was rich in smectite clay minerals, which have an affinity for attracting or adsorbing water molecules. Also, the river was in flood stage, with velocities of 2.5 meters per second (m/s) and depths of less than 3 meters (m). Both of these factors may have enhanced OPA formation. The laboratory experiments, which were run at water temperatures of 20–22 degrees Celsius

323 (oC), showed that the amount of OPA formation increased with increasing salinity, which is known to cause flocculation of clays. With a combination of fresh OSSA II oil (diluted with a kerosene-range petroleum product), Rio Desaguadero sediment, and solutions using natural seawater and dilution to obtain salinities of 0.35, 0.7, 1.2, 3.5, and 35 parts per thousand (ppt), OPA formation increased with increasing salinity, and at 35 ppt almost all of the oil was taken up in OPA formation (fig. 4). In contrast, a minimal amount of oil was present as OPAs with fresh oil, distilled water, and river sediment, but when the brackish river water was used (1.5 ppt salinity), about 25 percent of the fresh oil formed OPAs.

Figure 3. Types of oil-particle aggregates: A, single and multiple droplet aggregate, B, solid aggregate of large, usually elongated mass of oil with interior particles (dashed blue circles), and C, flake aggregate of thin membranes of clay aggregates that incorporate oil and fold up (modified from Stoffyn-Egli and Lee, 2002). Blue color represents particles and yellow represents oil.

100 90 80 70 OPA 60 in Rio Desaguadero water oil 50 (Lee and others, 2002) 40 Seawater (Lee and

Percent 30 others, 2002) 20 10 0 0.01 0.1 1 10 100 Salinity (parts per thousand)

Figure 4. Salinity in relation to the percent of oil in oil-particle aggregates (OPAs) from laboratory shaker tests of diluted heavy crude oil form the 2000 OSSA II spill into the Rio Desaguadero in Bolivia and its smectite-rich sediment. (Graph replotted from Lee and others, 2002).

324 Natural formation and submergence of OPAs occurred in the freshwater, low-gradient environment of the Kalamazoo River after the 2010 Enbridge Line 6B pipeline spill of diluted bitumen into Talmadge Creek, a tributary of the Kalamazoo River near Marshall, Michigan (Dollhopf and others, 2014). The diluent of natural gas condensate (which is used as a solvent in the mixture) volatilized and submergence of the bitumen occurred within a few weeks following the spill (Dollhopf and Durno, 2011; Lee and others, 2012). On the basis of laboratory tests of Cold Lake Blend by Belore (2010) and in an outdoor flume by King and others (2014), the density of the spilled bitumen (mainly Cold Lake Blend) was likely between 0.93 and 0.936 grams per cubic centimeter (g/cm3) with diluent and 0.981 g/cm3 after the diluent evaporated. The dynamic viscosity of the Cold Lake Blend in the Belore (2010) laboratory tests ranged from approximately 400 centipoise (cP) with diluent to more than 14,500 cP after evaporation of the diluent. Both the works of Belore (2010) and King and others (2014) were conducted at approximately 15oC. Some features of the Kalamazoo River likely enhanced the formation, resuspension, and deposition of OPAs. The Enbridge Line 6B pipeline release happened during a flood on the Kalamazoo River with an exceedance probability of 4 percent (for example, a 25-year event) (Hoard and others, 2010) and with a mean velocity of about 1.1 m/s and a mean depth of 1.2 m near the USGS streamgage at Marshall, Michigan (04103500). On the basis of later measurements of suspended sediment at the Marshall streamgage, it can be inferred that, at the time of the spill, the river had relatively low suspended sediment concentrations (less than 100 mg/L) (fig. 5) with suspended particle sizes mainly in the silt-sized range (65–75 percent) (fig. 6). Water temperatures were warm, in the range of 23–25 o C (Stephen Hamilton, Michigan State University, written commun., 2014). Floodwater increased turbulence in river flows and increased the presence of suspended particulate matter. Additional mixing from flows over two dams may also have played a role, although OPAs and submerged oil accumulated in the first 5 kilometers (km) of river length, between the spill source and the first dam. Aggressive sediment agitation techniques (raking, flushing, aeration, and skimming the river bottom physically or with water jets) were conducted in 2011 to liberate submerged oil as recoverable sheen in 2011 (Enbridge Energy L.P., 2011b; Dollhopf and others, 2014), potentially contributing to further OPA formation and transport of OPAs to downstream reaches (Lee and others, 2012). Lee and others (2012) found that oil from the pipeline spill readily formed OPAs when mixed with Kalamazoo River sediment in laboratory tests (fig. 7). Assuming that the fraction of spilled oil not recovered by conventional techniques was lost to submergence, the bitumen that submerged in the Kalamazoo River was greater than 300,000 liters, which is around 10 percent of the spilled oil. This is based on the Enbridge Energy, L.P., reported spilled amount of 3.2 million liters and recovered amount of 2.9 million liters after the first year (U.S. Environmental Protection Agency, 2011). Although the volume of oil released and remaining in the river are not finalized at the time of this writing (January 2015), the estimated percentage is similar to that found by Lee and others (2002) for distilled water and brackish water of the Rio Desaguadero (fig. 4). Even though OPA formation in freshwater may be less than that in seawater, the persistent residual submerged oil and oiled sediment in the Kalamazoo River resulted in a protracted cleanup that ultimately required dredging and has accounted for a major share of the cleanup costs, which have surpassed $1.2 billion (Dollhopf and others, 2014).

325 100

in 90 Kalamazoo River at Marshall, 80 Michigan; streamgage 04103500 70 liter 60 Kalamazoo River near Battle concentration,

per Creek, Michigan; streamgage

50 04105500 40 sediment

millgrams 30 20 10 Suspended 0 0 20406080100 Streamflow, in cubic meters per second

Figure 5. Streamflow in relation to suspended sediment concentration in the Kalamazoo River, Michigan, August 2012-March 2014. Streamflow during the pipeline release on July 28, 2010, during a receding flood event with a 4- percent exceedance probability, was about 34 cubic meters per second (m3/s) at Marshall, Michigan and 85 m3/s near Battle Creek, Michigan. The Kalamazoo River is generally a suspended sediment supply -limited system, shown by the overall low concentrations over the entire flow range at both streamgages.

100 90

80 Kalamazoo River at 70 Marshall, Michigan, streamgage 04103500 60

percentage Kalamazoo River near Battle 50 Creek, Michigan, 40 streamgage 04105500 30 Cumulative 20 10 0 1 10 100 1000 Particle size, in microns

Figure 6. Cumulative particle-size distribution for suspended sediment collected April 22, 2013, during flows of 32 cubic meters per second (m3/s) at Marshall, Michigan, and 85 m3/s near Battle Creek, Michigan. Most of the suspended sediment was in the silt-size class—about 65 percent at Marshall and 75 percent near Battle Creek. Samples analyzed with portable laser in-situ scattering and transmissiometry.

326

Figure 7. Kalamazoo River sediment spiked with weathered source oil after 48 hours, under ultraviolet- epifluorescence microscopy at 320 times magnification. (from Lee and others, 2012).

Some additional features of the Kalamazoo River may have been important factors in OPA formation, transport, and deposition. The floodplain of the Kalamazoo River has abundant wetlands, thus suspended and bottom sediments have relatively high organic matter content, on the order of 20 percent or more. The river is wide (width/depth ratios of 40) and has an average gradient of 0.06 percent in the spill-affected reach. Deposition of OPAs occurred along channel margins, backwaters, side channels, and oxbows, and in impoundments throughout the entire 61- km stretch of the river affected by the oil spill. Surveys over time provided evidence for resuspension and resettling of OPAs in downstream areas, presumably during post spill floods (Dollhopf and others, 2014).

Oil-Particle Aggregates as a Natural Physical Dispersant Enhancing physical dispersion of spilled oil through the addition of particulate matter is one of several techniques that have been used for spill cleanup and to prevent oiling of marine coastal areas (Zhang and others, 2010). The oil would be dispersed into small droplets by turbulent mixing from waves or currents and subsequently mixed with mineral and organic particles in the water column. The aggregation of the oil with particles to form OPAs and their subsequent physical dispersion by natural processes would reduce the bioavailability and toxicity of the residual oil to aquatic organisms in the vicinity of the spill. Furthermore, as the activity of oil degrading bacteria is focused at the oil water interface, the formation of small oil droplets enhances microbial biodegradation (Lee and others, 2002; 2009). Petroleum hydrocarbons are not new to the environment because of natural sources such as seepage from oil bearing rocks and biological production by plants and animals. Thus, indigenous oil degrading bacteria are readily available in many aquatic ecosystems (Atlas and Hazen, 2011). OPA formation has been reported to be a significant contributor to the natural cleansing mechanisms observed during the Baffin Island Oil Spill (BIOS) project and recovery following the Exxon Valdez oil spill incident in 1989 (Bragg and Owens, 1994; Bragg and Yang, 1995). Studies have shown that microorganisms (namely bacteria and archea) within ocean floor sediments in the proximity of natural oil seep sites, such as Scott Inlet, Baffin Island, and the Gulf of Mexico have adapted to utilize oil droplets as a carbon and energy source (Grant and others, 1986; Atlas and Hazen, 2011). The current scientific consensus is that a considerable portion of the oil spilled in the Gulf of Mexico from the Deepwater Horizon incident has been degraded by indigenous bacteria (Atlas and Hazen, 2011; National Research Council, 2013). Considering the magnitude of the spill, Edward Owens (Polaris Applied Science, Inc., oral commun., 2014) has hypothesized that the volume of oil affecting the Gulf of Mexico shoreline was much less than expected, because of the interaction of dispersed oil and surface oil slicks with the naturally high concentration of particulate suspended material near the coast.

327 Boufadel et al. (2014) conducted a comparison between the DWH and the Exxon Valdez spills, and based on calibrated modeling, they estimated that around the same mass of oil reached the shorelines from the two spills (around 20,000 tons). But the percentages to the total mass of oil were around 5 percent and 50 percent for the DWH and Exxon Valdez, respectively. For the 1989 Exxon Valdez oil spill, much of the oil along shorelines was depleted (Atlas and Hazen, 2011), except in some beaches with anoxic conditions (Boufadel and others, 2010; Li and Boufadel, 2010). With natural particulate matter readily available in a coastal marine or lacustrine environment, and especially along river systems, OPA formation must be considered a natural process that enhances the physical dispersion of oil. Indeed, expanding on this hypothesis, Lee and others (2009) suggested that active enhancement of OPA production as a “physical” means to promote oil dispersion, could be an alternative to the use of chemical dispersants that may be potentially toxic in their own right. However, prescribed sinking of spilled oil, or the use of sinking agents, is currently prohibited by U.S. Environmental Protection Agency (EPA) because of the potential risks of acute and chronic toxic effects on benthic organisms and possibly less biodegradation once the oil is deposited (U.S. Environmental Protection Agency, 1993; also 40 C.F.R. §§ 300.310(b), 300.910(e)). Laboratory experiments have expanded on the knowledge base of the manner in which OPAs facilitate physical dispersion in both seawater and freshwater. For seawater, simulations of coastal environments by Li and others (2007) in wave tank experiments found that chemical dispersants and the addition of fine mineral particles, alone and in combination, enhanced the dispersion of light crude oil in the water column and increased the number of OPAs formed. Similarly, results from wave tank studies of light crude oil with seawater (Lee and others, 2008) indicate that chemical dispersants enhanced OPA formation by transferring oil from a slick floating on the water surface into oil droplets in the water column that more easily interacted with suspended particles. Zhang and others (2010) considered three crude oils (Mesa, Alaska North Slope, and Heidrun) with specific gravities from 0.8746 to 0.9058 g/cm3 (at 22oC). They conducted laboratory experiments in seawater to evaluate the combined effects of three factors—chemical dispersants, mixing energy, and mineral types—on OPA formation. They also discussed the usage of OPA as a response technique. They found that hydrophobicity (aversion toward water), particle size, and specific surface area of minerals are key factors, which is consistent with other findings in the literature. Slightly hydrophobic particles enhanced formation of OPAs by promoting the attachment of mineral particles to oil. However, highly hydrophobic particles clumped together and did not interact with the oil. Therefore, there was an optimum range of hydrophobicity for maximum OPA formation. If minerals were hydrophilic, the OPAs were generally spherical. Hydrophobic minerals formed irregularly shaped OPAs. The OPAs were larger for hydrophobic particles than for hydrophilic minerals. Chemical dispersants when added became the overriding factor affecting OPA formation because of their stabilizing effect on oil droplets. High mixing energy enhanced dispersion of oil into the water column to form droplets and small-sized OPAs. The Zhang and others (2010) study concluded that in areas of low mixing energy, a chemical dispersant might be needed with a co-application of fine mineral particles to form OPAs. Laboratory studies in freshwater using automated shaker tests of mixtures of kaolinite clay-sized particles and heavy and intermediate fuel oils (viscosities of 3,900 and 1,350 cP, respectively) were conducted by Perez and others (2014) to simulate of the interaction of oil slicks and suspended sediment in steep, turbulent rivers. Using a spectrophotometer for oil measurement, the amount of oil entrained by sediment was observed to be moderate for heavy fuel oils at wave heights of 2.5 and 7 cm and kaolinite concentrations of up to 16,000 parts per million (ppm). They predicted that a surface slick of 1,000 kg of IFO across a river width of 10 m would result in about 8 percent of the oil entrained in OPA over a 1 km length. However, there was a large variability in the data, and thus these results still need

328 confirmation of scaling factors using either large scale experiments or with computational fluid dynamics models. Khelifa and others (2005b) and Niu and others (2010, 2011) developed predictive models to estimate the contribution of OPAs to the dispersal of spilled oil and their potential to cause secondary detrimental effects associated with physical inhibition (that is smothering of benthic organisms) or toxicity. Factors considered in these models include the calculation of the maximum size of droplets, prediction of oil droplet formation from a slick, prediction of sediment aggregate formation, and the calculation of the density of the resultant oil-sediment aggregate. Inputs to the models include environmental conditions, oil properties, and concentration and particle-size distribution of suspended sediment. In a model sensitivity analysis of five crude oils (Hibernia, Louisiana, Prudhoe Bay, Arabian Light, and Alaska North Slope), with a range of densities from 25 to 37 degrees API and dynamic viscosities from 8 to 68 cP, the kinetic energy dissipation rate was varied from 10-3 to 102 square meters per cubic second (m2/s3) with a sediment particle size of 3 micrometers (m) and concentration of 250 mg/L; the study showed that OPA formation is strongly dependent on the oil-water interfacial tension and kinetic energy dissipation rate. The OPA contribution to oil dispersion increased when energy dissipation rates were about 1 watt/kg or higher. High turbulence in surf zones or in rivers should therefore be conducive to higher rates of OPA formation.

Transport and Fate of Oil-Particle Aggregates As described in the section “Oil-Particle Aggregates as a Natural Physical Dispersant”, formation of OPAs changes the fate and transport of oil by potentially changing its rate of horizontal and vertical transport, and biodegradation and levels of bioavailability, which ultimately influence the ecological effects of OPAs. Because of the range of variance in physical, chemical, and biological conditions between sites (for example, types of suspended particulate organic/inorganic material, type of oil, mixing energy) differences in the transport and fate have been observed between marine and freshwater environments. In the coastal marine environment, the formation of OPAs has been found to improve removal of stranded oil from low-energy intertidal environments and is considered to be a natural self-cleansing process that enhances recovery rates following a spill (Lee, 2002). In contrast, in lowland rivers with gentle gradients, naturally formed OPAs can lengthen oil spill cleanup times and require deployment of less conventional and more costly sediment remedial measures (Dollhopf and others, 2014; Gouvernement du Quebec, Depot Legal, 2014). For example, in the 2010 Enbridge Line 6B spill of diluted bitumen into the Kalamazoo River, approximately 100 hectares (250 acres) of oiled sediment remaining in impounded sections of the river was removed by dredging and excavation during 2013–14, in response to persistent sheening problems (Dollhopf and Durno, 2011; Dollhopf and others, 2014). Oil globules and OPAs of various sizes, up to a few mm in size, were resuspended during floods, released upon mechanical agitation or physical disturbance of the sediment, and liberated by gas bubbles rising to the surface in a process called ebullition (as happens naturally when methane is produced in freshwater sediments) (fig. 1). Similarly, following the Lac-Mégantic light crude train spill in the Chaudière River, Quebec, Canada, oiled sediments created a challenge to clean-up operations downstream from the spill site over a 30 km reach of the Chaudière River (Gouvernement du Quebec, Depot Legal, 2014). Modeling the transport and fate of OPAs in riverine systems requires integration of hydrodynamic, sediment transport, and contaminant fate and transport models (Dollhopf and others, 2014; Niu and others, 2010, 2011), while employing some of the same guidelines used for developing conceptual and mathematical models of fate and transport of contaminated sediment at hazardous waste sites (U.S. Environmental Protection Agency, 2005). Simplified fate and transport studies were done for

329 the 2010 Kalamazoo River spill by considering OPAs in a steady state of physical properties, including size, concentration, specific gravity, erosion rates, and settling velocities, as well as hydrocarbon content (Dollhopf and others, 2014). More complex models, accounting for advection/diffusion, settling, resuspension, and breakup characteristics, are being considered for freshwater (riverine and deep-water settings) and marine environments (Lee and others, 2011a; Niu and others, 2011, 2014) and can be built from simpler models that simulate oil slicks (Weaver, 2004). For the Enbridge Line 6B diluted bitumen spill into the Kalamazoo River, a 2-dimensional (2-D) Environmental Fluid Dynamics Code (EFDC) model was constructed to determine areas of the river prone to resuspension and deposition of submerged oil and oiled sediment under different flow conditions (Hamrick, 1992; Enbridge Energy L.P., 2012a; Dollhopf and others, 2014). Initially, OPAs were assumed to behave similarly to silt-sized particles because the OPAs accumulated in depositional areas and impounded sections of the river with organic- and silt-rich soft sediment. To account for differences in behavior, the model was updated with a new algorithm for OPAs that was incorporated into the sediment transport code of the SNL-EFDC model—a modified version of the original EFDC code developed and maintained by Sandia National Laboratory (James and others, 2005; Thanh and others, 2008). This version of the EFDC model incorporates a custom sediment transport sub-model that is based on the SEDZLJ model algorithms developed by Craig Jones and Wilbert Lick at the University of California – Santa Barbara (Jones and Lick, 2001). The OPA algorithm includes particle classes for representing two types of OPA—a sediment coating on an oil droplet and oil droplets in a particle- dominated aggregate. This algorithm and its application to EFDC and SEDZLJ are under development by the U.S. Army Corps of Engineers Engineering Research and Development Center at the time of this writing (January 2015). Hydrodynamic model results of velocity and horizontal bed shear stress from 2- D and 3-dimensional (3-D) EFDC model simulations of various flows and containment situations on the Kalamazoo River helped to target areas of the 61 km of oil-affected Kalamazoo River where submerged oil and OPAs were subject to resuspension and downriver migration during high flows (Dollhopf and others, 2014; Enbridge Energy L.P., 2012a). Erosion rates of soft sediment became an important parameter in the modeling because these areas tended to have moderate/heavy oiling, and some of the areas in impoundments switched from depositional to erosional during high flows. The soft sediment was cohesive, and onsite sedflume tests were performed by the U.S. Army Corps of Engineers to parameterize the sediment transport and OPA algorithms in the EFDC models (Perkey and others, 2014). Critical shear stresses of soft sediment with moderate/heavy oiling ranged from 0.1 Pascals (Pa) at the surface to 1.0 Pa at 20 cm beneath the surface (Perkey and others, 2014). Another approach for fate and transport modeling in rivers is a Lagrangian approach, also known as particle tracking. This approach has been used by the Ven Te Chow Hydrosystems Laboratory at the University of Illinois for the Kalamazoo River and builds on the existing EFDC-based hydrodynamic model. For impoundments with accumulations of thick fine-grained sediment, the process of bubble formation and release from sediments (ebullition) is likely to be an important mechanism for resuspending OPAs in the water column and releasing oil as sheen on the water surface. Spontaneous releases of oil globules and floating OPAs have been observed regularly in the impounded sections of the Kalamazoo River during 2011–14, resulting in oil sheens at the water surface (Dollhopf and others, 2014). The impoundments had generally 0.5 to 4-m thick accumulations of fine-grained organic-rich sediment which, under anaerobic conditions, allows for bacterial generation of methane (McLinn and Stolzenburg, 2009). These bubbles could rise based on their size and buoyancy, and their release from the sediments could be enhanced by disturbances such as fish and boats movement, falling water levels,

330 and dropping barometric pressures. Methane production and ebullition are enhanced by warm summer water temperatures (as high as 28oC) that promote bacterial activity and reduce the solubility of methane in sediment porewaters.

Ecological Risk and Toxicity of OPAs and Oiled Sediment The aquatic toxicity of OPAs is of interest, whether the OPAs formed through addition of minerals as a dispersant or from association of oil droplets with suspended inorganic and organic particles naturally present in a water body after the spill. Once oil droplets are aggregated with particles and submerge, associated contaminants, especially polycyclic aromatic hydrocarbons (PAHs), become a potential problem to suspended and benthic aquatic organisms, including plankton, zooplankton, invertebrates, mussels, and clams, and any higher level organisms that consume them (Long and others, 1998; MacDonald and others, 2000; Passow and others, 2012; Almeda and others, 2013). On the other hand, as illustrated during field studies, oil bound up in OPAs may be diluted to below toxicity threshold limits (Lee and others, 2003b) and may become more available for biodegradation (Lee and others, 1997; Lee and Merlin, 1999). An important ecological consideration is closely linked to the turbulent energy of the environment with more risk associated with submerged OPAs in depositional or low energy environments. Freshwater depositional environments are common in low-gradient rivers, river mouths and harbors, impoundments, ponds, lakes, and wetlands. Low-energy marine environments are likely similar (Niu and others, 2011). Thus, a major factor in ecological risk is whether the OPAs are physically diluted in suspension or concentrated in deposition. The added context of water depth and the geographic extent are also important. However, oil droplets and OPAs have a size range similar to that of planktonic food sources for zooplankton, and biogeochemical processing by organisms could increase or decrease toxicity within different components of the aquatic food web (Passow and others, 2012; Almeda and others, 2013). Data are sparse on the specific toxicity of OPAs in marine and freshwater environments but there are a few studies to draw upon. Qualitative analysis indicated that residual oil was in a highly biodegraded (that is, less toxic) state in suspended particulate material associated with OPA formation from surf washing operations during the Sea Empress Spill in the United Kingdom (Lee and others, 1997). In terms of the application of bioassays, samples recovered from sediment traps deployed in the immediate area off an experimentally oiled beach site in Svalbard, Norway, remediated by the enhancement of OPA formation by surf-washing, were found to be within Environment Canada’s acceptable regulatory limits for the disposal of dredge spoils (Lee and others, 2003a; Lee and others, 2003b). For freshwater riverine environments, limited acute toxicity testing was done for oiled sediment in the Kalamazoo River after the Line 6B oil release (Bejarano and others, 2012). As part of a Net Environmental Benefits Analysis, effects on aquatic organisms from weathered oil were assessed in laboratory acute toxicity studies of seven sediment samples collected from oil-affected backwater habitats along the Kalamazoo River in February 2012, about 19 months post-spill (Bejarano and others, 2012). Ten-day whole sediment toxicity tests using Chironomus dilutus (a species of midge) and Hyalella azteca (an amphipod crustacean) were performed by the Great Lakes Environmental Center, Inc., and included survival, growth and biomass as the toxicity endpoints (Great Lakes Environmental Center, 2012). Results from the toxicity tests indicated that Chironomus dilutus were more sensitive to oiled sediment (and presumably OPAs) than Hyalella azteca but that all samples exceeded the minimum survival (70 percent) and growth (0.48 mg ash-free dry weight at test termination) criteria for acceptable controls for the C. dilutus tests (Great Lakes Environmental Center, 2012).

331 Chemical analyses (PAHs, total extractable hydrocarbons) and sediment characterization (total organic carbon and sediment composition) were performed on a subset of the above described Kalamazoo River sediment toxicity samples (Bejarano and others, 2012; Great Lakes Environmental Center, 2012). Potential adverse acute and chronic effects on benthic organisms were evaluated using the Equilibrium Sediment Benchmark Toxic Unit Approach (U.S. Environmental Protection Agency, 2003) for sediment PAH data from the same samples used for the bioassays. The results from these analyses indicate that sediment from two heavily oiled sites and one lightly oiled site may pose acute and chronic risks to benthic fauna (Chironomus dilutus and Hyalella azteca). However, further analyses of the toxicity results in the context of several other sediment characteristics (chemical and physical) showed that unrelated variables, such as percentage of silt in the sediment sample, may have affected survival (table 1). On the basis of the weight of evidence approach and additional risk metrics, it is possible that residual oil from the Enbridge Line 6B oil spill at two heavily and one lightly oiled area may pose some risks to benthic receptors. Chronic toxicity effects from the Enbridge Line 6B residual oil remain unknown at the time of this writing (January 2015).

Table 1. Nonparametric Spearman's Rho correlation coefficients between acute sediment toxicity test results and analytical variables using Chironomus dilutus and Hyalella azteca. Highlighted cells indicate a statistically significant correlation at α=0.05. Negative correlation coefficients indicate that survival, growth, and biomass were reduced in sediment with higher concentrations of low molecular weight polyaromatic hydrocarbons (PAHs) and total extractable hydrocarbons (TEHs) (from Bejarano and others, 2012). [LMW, low molecular weight; HMW, high molecular weight; µg/kg, micrograms per kilogram; mg/kg, milligrams per kilogram; TPAH, total polyaromatic hydrocarbons; TOC, total organic carbon; %, percent]

Chironomus dilutus Hyalella azteca Variables Survival Growth Biomass Survival Growth Biomass Sum LMW-PAH (µg/kg)* -0.53 -0.47 -0.51 -0.44 -0.18 -0.17 Sum HMW-PAH (µg/kg)* -0.45 -0.36 -0.40 -0.36 -0.20 -0.17 TPAH (µg/kg) -0.49 -0.43 -0.47 -0.39 -0.22 -0.21 TEH (mg/kg)** -0.52 -0.53 -0.54 -0.47 -0.29 -0.39 %TOC -0.67 -0.35 -0.42 0.04 -0.65 -0.65 % Gravel -0.08 0.02 0.03 0.01 0.07 0.18 % Sand 0.56 0.40 0.41 0.23 0.41 0.54 % Silt -0.54 -0.52 -0.52 -0.36 -0.44 -0.55 % Clay -0.53 -0.21 -0.25 -0.13 -0.41 -0.53 *Low molecular weight (LMW) PAHs include Naphthalene to Benzo(b)fluorine (38 analytes), whereas high molecular weight (HMW) PAHs include Fluoranthene to Benzo[g,h,i]perylene (26 analytes). ** Total Extractable Hydrocarbons (C9-C44; TEH).

Summaries of the physical effects from burial of benthic organisms by deposited OPAs can be found in available literature on burial effects from agitation dredging (Chapman, 2012) and on sediment deposition in streams (Waters, 1995). Aquatic organisms most likely affected by burial are fish eggs, larvae, and fry; sessile filter feeders such as mussels; and macrophytes (Chapman, 2012; Morton, 1977; and studies cited in Kaplan and others, 1974; Erftemeijer and Lewis, 2006). For marine benthic organisms, recommendations have been made to limit disposed sediment to 15 cm, but species survival is highly variable, with some species destroyed by as little as 5 cm (OSPAR, 2008) and fish eggs by as little as a few mm (Berry and others 2003; see other citations in Chapman, 2012). Deposition effects vary by species requirements, extent and spatial connection of habitat types, and size of the waterbody.

332 Effects of Ice in Northern Climates on OPA Formation and Spill Response Most of the cold-condition studies that have been conducted have looked at the behavior of spilled oil and not how the oil interacted with particles to form OPAs (Fingas and Hollebone, 2003; Lee and others, 2011b; Wang and others, 2013) under controlled laboratory and pilot-scale tank studies. Lee and others (2011b) summarize findings from laboratory studies funded by the Canadian Coast Guard to elucidate the potential significance of OPA formation under cold climatic conditions that included ice (Cloutier and others, 2005; Khelifa and others, 2005c). These studies showed that OPAs can form quickly with strong turbulence, within the first 10 minutes of a spill. Within 40 minutes, most of the oil was converted to OPAs. Brackish water (18 ppt) with slush and broken ice had similar results with strong turbulence. Most particles were less than 1 mm in diameter and about 50 percent of the oil was physically dispersed within 30 minutes (Blouin and Lee, 2007; Cloutier and Doyan, 2008). Field trials in the St. Lawrence River estuary near Matane, Quebec were conducted during the winter to further evaluate the feasibility of enhanced OPA formation as a spill response countermeasure in ice-infested waters (Lee, Li and others, 2011b). An icebreaker’s propeller was used to generate strong turbulence during this exercise to facilitate the formation of OPAs with experimentally released fuel oil sprayed with slurry of fine-grained chalk using the fire hose system on board the ship. Visual observations and results of laser particle size analysis (LISST) indicate that fuel oil physically dispersed as OPA into the water column by this experiment did not readily reform a surface slick. Without the addition of mineral fines and consequential OPA formation, the fuel oil resurfaced within minutes and was difficult to recover because of interference by large ice blocks. Half of the total petroleum hydrocarbons in recovered samples biodegraded after 56 days incubation at a temperature of 0.5 oC, which is most likely because of microbes in the water that were well adapted to their surrounding environment, despite the low temperatures. Although not specific to OPAs, Belore (2010) provided a detailed description of simulations of oil and condensate spills over a range of expected temperatures at the marine trans-shipment terminal and confined channel assessment area (CCAA) of concern for syncrude synthetic light oil, Condensate Blend (CRW), Cold Lake bitumen diluted with condensate, and MacKay River heavy bitumen diluted with Suncor synthetic light oil (MKH). The simulations included a hypothetical marine terminal spill and three hypothetical tanker spills for Emilia Island, Principe Channel, and Wright Sounds for spring, summer, fall, and winter conditions. The tests did not consider interactions between oil and particulate matter, but they provided a thorough set of physical properties and behaviors for these oils including density, viscosity, interfacial tension, pour point, flash point, evaporation, emulsion formation, and oil adhesion. The formation of OPAs and its effect on promoting dispersion of spilled residual oil may vary because of changes in the properties of the oil, water, particle type, temperature, and extent of ice cover. For example, the viscosity of diluted Cold Lake Blend ranges over an order of magnitude from summer water temperatures (for example, 393.2 square millimeters per second (mm2/s) at 15 oC) to winter water temperatures (1437.8 mm2/s at 1 oC) (Belore, 2010). Biodegradation and toxicity need more research because potential exposure and biological activity are affected by temperature. Recent results by McFarlin and others (2014) found that oil degrading microorganisms were present in surface, middle, and deep water samples from the Arctic Ocean and that oil biodegradation potential exists for offshore Arctic environments; however, this study did not specifically look at biodegradation of OPAs. Researchers from the Institute of Northern Engineering–University of Alaska Fairbanks and NewFields LLC have conducted several studies of the toxicity of physically and chemically dispersed

333 oil in arctic environments (McFarlin and others, 2011) that provide some insights into OPA toxicity. In one study, the toxicity of physically versus chemically dispersed oil to selected arctic species representative of the Beaufort and Chukchi Seas was examined at typical arctic water temperatures of 2 oC (McFarlin and others, 2011). Using fresh Alaska North Slope oil, physically dispersed under increased mixing energy in a laboratory, spiked exposure toxicity testing was performed for three arctic species: Calanus glacialis (copepod), Boreogadus saida (arctic cod), and Myoxocephalus sp. (sculpin). When subjected to physically dispersed oil, mean lethal concentration (LC50) values were lower by 3.3 and 3.7 mg/L total petroleum hydrocarbons (TPH) for arctic cod and early season copepods, than the corresponding values in the presence of chemically dispersed oil of 55 mg/L and 22 mg/L, respectively. With the exception of this case, toxicity effects for arctic species from physically dispersed oil were found to be generally no better or worse than for temperate species and warm water temperatures. In conclusion, the toxicity of suspended and deposited OPAs requires further investigation. OPAs that submerge and accumulate in depositional areas have chemical and physical ecological risks associated with them, either because of toxicity from high concentrations of hydrocarbons, or physical damage, which can result from the smothering and burial of benthic organisms. The appropriate application of laboratory tests to different geographic areas with a range of water depths and water currents needs further consideration.

Operational Considerations As described in preceding sections, oil can submerge, and OPAs can form rapidly after a spill of light and heavy crude oils, given the right environmental conditions. Key environmental factors that increase the probability of submergence for bitumen spills include (Silliman, 2014):  Low salinity resulting in relatively low water density,  Particles that have relatively high roughness in their surface area or are porous,  High turbidity or high suspended sediment concentrations,  Exposure to sunlight (UV radiation),  Strong currents and mixing energy, and  High temperatures. Silliman (2014) concludes that if one of these factors is present, then emergency response personnel should equip themselves with response tactics for submerged oil. The following sections describe the state of the science for detection, containment, and recovery of submerged oil, especially submerged OPAs.

Detection OPAs can be detected by direct observation using specialized microscopy techniques that enable the visualization of oil. On the basis of the strong natural fluorescence of aromatic hydrocarbons and chlorophyll when excited by UV light, Lee and others (1985) devised a bright field transmitted light/UV epi-fluorescence illumination technique that enabled the observation of interaction between chemically dispersed oil droplets and phytoplankton. This technique was subsequently refined and used in a routine manner to characterize and quantify OPAs (Stoffyn-Egli and Lee, 2002; Lee and others, 2003a; Ma and others, 2008). Detailed investigations of the surface and internal structure of OPAs have been conducted by the application of confocal laser scanning microscopy (Stoffyn-Egli and Lee, 2002; Zhang and others, 2010; Wang and others, 2011). Environmental scanning electron microscopy (ESEM) was used to produce high resolution 3-D topographical images of OPA surfaces in their natural state to

334 confirm the existence of “droplet” OPA composed of oil droplets that were stabilized by mineral fines on their surface (Stoffyn-Egli and Lee, 2002). The U.S. Coast Guard’s Research and Development Center recently specified requirements for submerged oil detection systems that included 80 percent detection probability, 1-m radius of detection, real-time results, reasonable setup time, accommodation of 1.5-m (5-ft) seas and 0.8 m/s (1.5 knot) currents, and coverage of 1.6 km2 (1 mi2) in a 12-hour shift (Hansen, 2014). Established on-site mass spectrometer systems to monitor oil in the water column may not be effective if the oil occurs as OPAs or has already settled out. This was the case in the oil-affected reach of the Kalamazoo River. Laser fluorometers offer the potential for detection of oil in oil-sediment mixtures and were successful in detecting residual oil in pore water in beach deposits along the coast of Taean, South Korea, following the Hebei Spirit oil spill (Kim and others, 2010). In addition, trials with submersible fluorometers with an excitation wavelength of 120-325 nanometers (nm) and emission wavelengths of 410-600 nm met with some success for the Kalamazoo River when the OPAs were in suspension during in situ tests of erosion characteristics of deposited oiled sediment from the 2010 Enbridge Line 6B oil release. A limited number of samples collected at the time the OPAs were in suspension were analyzed at the USGS Wisconsin Water Science Center for both absorbance and fluorescence excitation-emission matrices using an Aqualog instrument (Peter Lenaker, U.S. Geological Survey, written commun., 2013). However, once the OPAs are submerged and mixed with bottom sediment the fluorescent properties of the oil are masked by other particles (Lee and others, 2012). As a result of these short-falls identified with current methodologies for oil spill detection in aquatic environments, Hansen (2014) stressed the importance of simultaneous use of multiple systems. Detection of OPAs by sonar can be difficult because the detectable sonar signature associated with pure oil is diminished when it is mixed with suspended and bottom sediments. Furthermore, particle size is a factor; oil associated with fine-grained (silt and clay-sized) organic-rich particles is difficult to detect by sonar (Hansen, 2014; authors’ experience on the Kalamazoo River cleanup). A novel sediment poling technique was developed for detecting and assessing the spatial distribution of submerged oil and oiled sediment in the Kalamazoo River; it was adapted from studies of contaminated sediment (David Richardson, Tetra Tech, Inc., oral commun., 2011). This became the primary submerged oil assessment tool used in the Kalamazoo River cleanup (Enbridge Energy, L.P., 2013a; Dollhopf and others, 2014). The sediment was agitated using a graduated aluminum pole with a 20-cm-diameter metal disc on the submerged end. If submerged oil was present in the sediment, the agitation action liberated oil from the sediment, allowing it to float to the water surface. The percent coverage of oil sheen and number of globules at the water’s surface within 1 m2 were observed and categorized as none, light, moderate, or heavy according to the field observation submerged oil flowchart (fig. 8). Thickness of soft sediment in depositional areas could be estimated by quantifying the difference between a 1- and 2-hand push of the graduated pole into the sediment. This procedure was used to map the relative concentration and extent of oiled areas to depositional areas of the river with soft sediment (silt and organic matter) accumulations. From 2010 to 2014, over 20,000 poling points were assessed throughout the affected river system. Global positioning system (GPS) coordinates and field observations were recorded upon poling and managed in a geographic information system, which allowed for detailed mapping of sheening as well as bathymetry, substrate, and flow velocity. Some limitations of the method include the need for water temperatures generally greater than 15oC for consistent categorization, velocities less than about 0.3 m/s (otherwise the sheen and globs are swept too quickly downstream), and water depths of generally less than 3 m. At greater water depths, it is unknown whether the liberated oil can rise the full distance to the water surface, resettles, or is transported downstream.

335 For the Kalamazoo River spill, multiple rounds of sediment cores were collected, and two types of in situ suspended sediment samplers were deployed for monitoring and assessment (Enbridge Energy, L.P., 2011a). About 25 in situ suspended sediment samplers (Phillips and others, 2000) were placed along the 61-km stretch of the affected river to collect any submerged oil and oiled sediment in suspension in the water column that was being transported over a range of flows including during cleanup activities. In addition, about 70 suspended sediment settling jars, also known as sediment traps (Thomas and others, 2007), were placed in depositional areas of the Kalamazoo River with little to no flow to collect particles that were recently in suspension and settled to the river bottom. These jars were similar to column sampling devices used in lacustrine environments for particulate resuspension and chemical fluxes with ponded water or multiple flow directions, such as wave action and longshore currents (Eadie and others, 1984; Murdoch and MacKnight, 1994). Initially in the cores, globs of submerged oil from the Kalamazoo River Line 6B spill fluoresced under UV light, making it possible to identify the presence and depth of oil in the cores. In 2012, 2 years after the spill, it became necessary to positively identify that the oil in river sediment originated from the spilled oil and not from a previous spill or other background hydrocarbons (Dollhopf and others, 2014). It was at this time that the oil in the sediment cores was becoming more difficult to visually quantify using UV-fluorescence because of dilution, dispersion, and quenching of oil droplets within the sediment matrix (Lee and others, 2012). It then became necessary to identify a chemical fingerprint of the Line 6B oil using forensic oil chemistry techniques that distinguished a signature of triaromatic stearene biomarkers in Line 6B oil that was unique from other oil contaminants present in Kalamazoo River sediment before the spill (G. Douglas, NewFields, written commun., 2014; Dollhopf and others, 2014). This technique was used to quantify the remaining Line 6B oil concentrations in all three types of sediment samples collected in the Kalamazoo River—sediment cores, in situ suspended sediment samples, and recently deposited samples.

Containment Conventional containment strategies for floating oil do not address the submergence, potential resuspension, and subsurface transport and redistribution of OPAs. OPAs will easily pass underneath a surface boom. Subsurface booms and silt curtains are more effective. An example of the equipment used to keep submerged oil and oiled sediment from the 2010 Line 6B spill from migrating farther downstream in the Kalamazoo River and into Morrow Lake is shown in figure 9A. The top curtain contains a boom similar to conventional setups to trap floating oil. The bottom curtain is meant to cut off bottom currents and promote deposition. This type of containment curtain is oriented at an angle to river currents to maximize settling of OPAs while limiting the chance for new areas of scour (fig. 9B) (Enbridge Energy, L.P., 2012b, 2013b).

336

Figure 8. Flowchart used for field observations of submerged oil during poling assessments. (from Enbridge, 2013a; %, percent; <, less than; >, greater than).

Recovery Recovery techniques for submerged oil and OPAs in freshwater and marine environments are still in the development phase. In a recent document on the fate and transport of potentially spilled oil associated with the proposed Energy East Pipeline Project from Alberta to New Brunswick the following recovery equipment and techniques were recommended by Energy East Pipeline Ltd. (2014): specialized nets, bottom booms, dams and underflow weir dams, dredging, manual recovery, and air injection. Challenges of oil recovery in cold climates, whether in marine or freshwater environments, involve accounting for the following variables: presence of ice, air and water temperatures, remote locations, and low solar radiation (Lee, Li and others, 2011a, b). Early on in spills, weighted sorbent materials can be dragged along the sea floor or draped along a river bottom to capture submerged oil and OPAs (Hansen, 2014; Enbridge Energy, L.P., 2010). Subsurface sorbent pom-poms (Pister and others, 2009) have been used for oil in suspension. Where oil in deposited OPAs remains at concentrations that cause concerns for benthic organisms or excessive sheening problems, dredging may be necessary (Dollhopf and others, 2014; Gouvernement du Quebec, Depot Legal, 2014). Enbridge used a combination of agitation toolbox techniques and sheen sweeps in contained areas of the Kalamazoo River, but sheening problems continued in depositional areas after these techniques were used, leading to the adoption of dredging as the final solution (Enbridge Energy, L.P., 2010; Dollhopf and others, 2014). Agitation toolbox techniques used on Kalamazoo River bottom sediment included mechanical agitation through raking, hand-held tillers, and chain drags, along with hydraulic agitation using hand-held water jets arranged as a single wand or on a rotating head (known as stingers) and vessel-mounted or dragged spreader bars with multiple water jets (Enbridge Energy L.P., 2011b). Conventional oil skimming techniques and sheen sweeps were used in response operations in the Kalamazoo River cleanup during periods of spontaneous release of oil globules from depositional areas of the river.

337 A

Figure 9. A, Diagram of subsurface curtain boom, and B, map of containment boom locations for submerged oil in Morrow Lake Delta from river miles 37.25 to 37.75, Kalamazoo River, Michigan related to the Enbridge Line 6B oil release. Blue arrows on map show general flow direction from right to left. Bold dashed lines are containment boom locations. Redrawn diagrams from Enbridge Energy, L.P. (2013b) containment permits.

338 Some of the methods that have been used for OPAs in suspension and on the bottom along shorelines include (Pister and others, 2009)  Weighted sorbent materials (for example, plastic pom-poms),  Dredging,  Natural attenuation,  Vacuum truck (limited by access),  Small portable submersible pumps,  Debris removal and wrack cleaning,  Sediment reworking,  Flooding, and  Low and high pressure flushing (ambient water) (may leave a significant quantity of oil that requires additional effort).

Techniques that were generally not recommended for marine shorelines and beaches include (Pister and others, 2009)  Offshore barriers and berms,  Mechanical oil removal,  Vegetation removal,  Low and high pressure flushing (hot water),  Steam cleaning,  Sand blasting,  Solidifiers,  Shoreline cleaning agents, and  Natural microbe seeding.

Unfortunately, there is not a simple operational endpoint for spill clean-up operations when it comes to residual OPAs. Remediation for each spill, whether in marine or freshwater environments, can benefit by the development of a Net Environmental Benefit Analysis, which weighs the benefits and drawbacks of leaving oil in place rather than causing further physical damage to aquatic habitats by aggressive removal techniques such as dredging or agitation (Efroymson and others, 2003; Rayburn and others, 2004; Bejarano and others, 2012). It cannot be automatically assumed that in all situations the oil concentrations in deposited OPAs will fall below toxic concentrations during a short period of time as a result of natural attenuation (dilution and biodegradation). Furthermore, it is important to note that the biodegradation of residual oil associated OPA is disadvantaged under anaerobic conditions that form after burial (Lee, 2000). However, depending on the spatial extent of the OPAs, the environmental setting, or presence of sensitive habitat, oil concentrations in OPAs may be adequately diluted to warrant no recovery and allow natural attenuation to happen (Lee and others, 2011a; Bejarano and others, 2012; Fitzpatrick and others, 2013).

339 Future Science Needs This report has summarized the state of knowledge regarding the formation of OPAs in natural waters, their eventual transport and fate, and considerations for cleanup of oil spills. The existing knowledge base is insufficient but rapidly growing with recent observations from large spills in freshwater and brackish riverine environments, additional laboratory tests and modeling, and studies of environmental effects of proposed new and expanding pipeline systems in North America carrying bitumen products (http://nas-sites.org/dilbit/). Specific science needs for submerged oil and OPAs are listed below.  Laboratory experiments of resuspension and breakup of OPAs.  Updated and new models and simulations of fate and transport of oil and OPAs in freshwater and cold climate environments with a range of oil and sediment types.  Quantitative monitoring and mapping of large areas of OPAs in water depths greater than 3 m.  Refinement of the Kalamazoo River poling technique with deployment of a fluorometer in the plume of sediment OPAs resuspended by the poling agitation.  Monitoring and assessment of transport and fate of spreading oil in ice and below ice.  Field trials in cold climates. More study is needed using realistic field trials or field observations of OPA formation during oil spills. Especially needed are studies of freshwater environments in and around the Great Lakes coastal environments and river mouths.  Investigation of potential effects on benthic invertebrate communities from residual oil and OPAs in depositional environments including burial and smothering as well as hydrocarbon toxicity in marine, freshwater, and cold-climate environments.  Investigation of OPA toxicity and physical effects on habitat. Not enough is known yet about the toxicity of OPAs, especially chronic toxicity and routes of exposure, and application of laboratory results to specific aquatic habitats. Data on the potential negative effects of augmented natural dispersion on burial and smothering of benthic organisms is needed, especially for freshwater environments.  Vulnerability analyses of critical habitats.  Incorporation of OPA properties into hydrodynamic and sediment transport models.  Post-spill monitoring and assessment techniques.  Operation endpoints—monitoring protocols to determine how much cleanup is enough and the manner in which natural attenuation may ameliorate effects in the future. Summary and Conclusions Studies of the formation of oil-particle aggregates (OPAs), and related behavior, fate, and toxicity in a wide variety of environments, including freshwater rivers, are of continued interest to researchers as transportation of light and heavy crude oils continues to increase across North America. This report contains an up-to-date review of the state of the science for OPAs from available literature, in terms of formation and stability, use as a physical dispersant, transport and fate, toxicity, behavior in cold climates, operational considerations, and future science needs. Although much is known about OPAs, there remains a good deal of science to be learned, especially in terms of laboratory experiments,

340 flume studies, toxicity and habitat effects, field exercises, and modeling at a range of spatial and temporal scales. Questions that were investigated for this report, along with brief answers developed from the available literature are listed here.  What is the effectiveness of adding particles to an oil spill for physical dispersion of oil as a spill countermeasure? The effectiveness varies with oil and particle properties and the salinity of water, but in general the addition of particles will almost always result in some OPA formation. The particles act to stabilize oil droplets and prevent them from re- coalescing at the water surface into an oil slick. Some potentially negative consequences occur when OPAs settle to the bottom possibly causing issues with protracted and increased cleanup costs, and loss of habitat to benthic organisms from smothering and burial. The toxicity of OPAs compared to oil droplets in the water column varies or is not known. Because of these negative effects, prescribed sinking of spilled oil, or the use of sinking agents, is prohibited by U.S. Environmental Protection Agency.  What is the long-term fate and transport of OPAs? Particle interactions tend to stabilize the oil droplets. OPAs can be resuspended when currents or mixing energy increases. OPAs can also be released to the water surface through the process of bubble formation and release from sediments (ebullition) for years following a spill.  What are the ecological implications of OPAs? Ecological implications have to do with whether the OPAs stay in suspension or settle out and the geographic extent and water depth of the aquatic habitat. If OPAs stay in suspension, then microbial degradation would be likely increased. If OPAs settle out, ecological effects would be likely from toxicity and physical smothering.  What are the operational considerations for recovery of OPAs? Containment and recovery of submerged OPAs or OPAs in riverine and marine environments require different techniques than those used for floating oil, and a familiarity with hydraulics of sediment transport is helpful.  Are there special circumstances for OPAs in cold climates? OPAs form readily in cold climates and the addition of particles likely acts as a physical dispersant similar to warm climates. The additional difficulty of using conventional skimming techniques for floating oil where there is broken ice makes physical dispersion a more attractive option. However, habitat loss, burial, smothering, and toxicity effects from submerged OPAs need to be considered, especially for shallow freshwater environments. References Cited Almeda, R., Wambaugh, Z., Wang, Z., Hyatt, C., Liu, Z., and Buskey, E.J., 2013, Interactions between zooplankton and crude oil: toxic effects and bioaccumulation of polycyclic aromatic hydrocarbons: PLoS ONE v. 8, no. 6, p. 1–21, accessed January 10, 2015, at http://www.plosone.org/article/fetchObject.action?uri=info%3Adoi%2F10.1371%2Fjournal.pone.006 7212&representation=PDF. American Petroleum Institute, 2011, Crude oil category, category assessment document: American Petroleum Institute Petroleum HPV Testing group, January 14, 2011, 108 p., accessed January 10, 2015, at http://www.epa.gov/oppt/chemrtk/pubs/summaries/crdoilct/c14858ca.pdf.

341 Andrews, A., 2014, Crude oil properties relevant to rail transport safety: in brief: Congressional Research Service, 7-5700, R43401, February 18, 2014, 10 p., accessed January 10, 2014, at https://www.hsdl.org/?view&did=751042. Association of American Railroads, 2014, Moving crude oil by rail: Association of American Railroads, September 2014, accessed January 10, 2015, at https://www.aar.org/BackgroundPapers/Crude%20Oil%20by%20Rail.pdf. Atlas, R.M., and Hazen, T.C., 2011, Oil biodegradation and bioremediation: a tale of the two worst spills in U.S. history: Environmental Science and Technology, v. 45, p. 6,709–6,715. Attanasi, E.D., and Meyer, R.F., 2007, Natural bitumen and extra-heavy oil, in Trinnaman, J., and Clarke, A., eds., 2007 Survey of Energy Resources: World Energy Council, p. 119–143. Bandara, U.C., Yapa, P.D., and Xie, H., 2011, Fate and transport of oil in sediment laden marine waters. Journal of Hydro-environment Research, v. 5, p. 145–156. Bejarano, A.C., Michel, J., and Williams, L., 2012, Net Environmental Benefit Analysis (NEBA) relative risk ranking conceptual design, Kalamazoo River system, Enbridge Line 6B Release: Enbridge Line 6B MP 608, Marshall, MI, August 8, 2012; document and appendixes; AR-0963. Belore, R., 2010, Properties and fate of hydrocarbons associated with hypothetical spills at the marine terminal and in the confined channel assessment area: Ottawa, Ontario, SL Ross Environmental Research Ltd., Technical Data Report, [variously paged]. Berry, W., Rubinstein, N., Melzian, B., and Hill, B., 2003, The biological effects of suspended and bedded sediment (SABS) in aquatic systems, a review: Internal Report. August 20, 2003: U.S. Environmental Protection Agency, Office of Research and Development, National Health and Environmental Effects Laboratory, Narragansett, RI, and Duluth, MN, accessed January 10, 2015, at http://water.epa.gov/scitech/swguidance/standards/criteria/aqlife/pollutants/sediment/upload/2004_08 _17_criteria_sediment_appendix1.pdf. Blouin, M., and Lee, K., 2007, Oil mineral aggregate formation in ice-infested waters: Proceedings of the 2007 International Oil & Ice Workshop. Minerals Management Service, Herndon, Va. Boufadel, M.C., Li, H., Suidan, M.T., and Venosa, A.D.. 2007, Tracer studies in a laboratory beach subjected to waves: Journal of Environmental Engineering v. 133, no. 7, p. 722–732. Boufadel, M.C., Nasab, A-A., Geng, X., Galt, J., and Torlapati, J., 2014, Simulation of the landfall of the Deepwater Horizon oil on the shorelines of the Gulf of Mexico: Environmental Science and Technology, v. 48, p. 9,496−9,505. Boufadel, M.C., Youness, S., Van Aken, B., Wrenn, B.A., and Lee, K., 2010, Nutrient and oxygen concentrations within the sediments of an Alaskan beach polluted with Exxon Valdez Oil Spill: Environmental Science and Technology, v. 44, p. 7,418–7,412. Bragg, J.R., and Owens, E.H., 1994, Clay–oil flocculation as a natural cleansing process following oil spills––Part 1: Studies of shoreline sediments and residues from past spills: Proceedings of the 17th Arctic and Marine Oilspill Program (AMOP) Technical Seminar, Environment Canada, Ottawa, Ont., p. 1–23. Bragg, J.R., and Yang, S.H., 1995, Clay-oil flocculation and its role in the natural cleansing in Prince William Sound following the Exxon Valdez oil Spill, in Wells, P.G., Butler, J.N., Hughes, J.S., eds., Exxon Valdez oil spill: fate and effects in Alaskan waters: Philadelphia, Pa., ASTM STP 1219, p. 178–214.

342 Center of Offshore Oil and Gas Environmental Research, 2004, The COOGER Update, March 2004, v. 1, no. 1, 6 p., accessed January 10, 2015, at http://www.dfo-mpo.gc.ca/science/coe-cde/cooger- crepge/newsletter/mar04_e.PDF. Chapman, J., 2012, Review of potential ecological effects of sediment agitation, Appendix B, in Bejarano, A.C., Michel, J., and Williams, L., Net Environmental Benefit Analysis (NEBA) relative risk ranking conceptual design, Kalamazoo River system, Enbridge Line 6B Release: August 8, 2012; Unpublished document and appendixes; AR-0963. Clark, M.M., 2009, Transport modeling for environmental engineers and scientists (2d ed.): Hoboken, New Jersey, John Wiley and Sons, Inc., 664 p. Cloutier, D., Amos, C.L., Hill, P.R., and Lee, K., 2003, Oil erosion in an annular flume by seawater of varying turbidities: A critical bed shear stress approach: Spill Science and Technology Bulletin, v. 8, p. 83–93. Cloutier, D., and Doyon, B., 2008, OMA Formation in Ice-Covered Brackish Waters: Large-Scale Experiments, in Davidson, W.F., Lee, K., and Cogswell, A., (eds.), Oil Spill Response: A Global Perspective: New York, Springer, p. 71–88. Cloutier, D., Gharbi, S., and Boule, M., 2005. On the oil-mineral aggregation process: A promising response technology in ice-infested waters: Proceedings of the 2005 International Oil Spill Conference, Miami Beach, FL, USA, p. 527-531. Coastal Response Research Center, 2007, Submerged oil—state of the practice and research needs: Durham, New Hampshire, Coastal Response Research Center, 29 p. + appendix. Committee for a Study of Pipeline Transportation of Diluted Bitumen, 2013, Effects of diluted bitumen on crude oil transmission pipelines: Washington, D.C., Transportation Research Board Special Report 311, 132 p. Crude Quality, Inc., 2014, Crudemonitor.ca., accessed January 10, 2015, at http://www.crudemonitor.ca/home.php. Delvigne, G.A.L., and Sweeney, C.E., 1988, Natural dispersion of oil: Oil and Chemical Pollution, v. 4, no. 4, p. 281-310. Doelling, P., Davis, A., Jellison, K., and Miles, S., 2014, Bakken crude oil spill, Barge E2MS 303, Lower Mississippi River, February 2014, National Oceanic and Atmospheric Administration, overview presentation, accessed January 10, 2015, at http://www.nrt.org/production/NRT/RRT3.nsf/Resources/May2014_pdf/$File/Bakken_Crude_Spill_E2M S303_Revised.pdf. Dollhopf, R.H., and Durno, M., 2011, Kalamazoo River Enbridge Pipeline Spill: Proceedings 2011 International Oil Spill Conference, accessed January 10, 2015, at http://ioscproceedings.org/doi/pdf/10.7901/2169-3358-2011-1-422. Dollhopf, R.H., Fitzpatrick, F.A., Kimble, J.W., Capone, D.M., Graan, T.P., Zelt, R.B., Johnson, R., 2014, Response to heavy, non-floating oil spilled in a Great Lakes river environment: a multiple- lines-of-evidence approach for submerged oil assessment and recovery: Proceedings, 2014 International Oil Spill Conference, Savannah, GA, May 7-9, 2014, p. 434–448, accessed January 10, 2015, at http://ioscproceedings.org/doi/pdf/10.7901/2169-3358-2014.1.434. Eadie, B.J., Chambers, R.L., Gardner, W.S., and Bell, G.L., 1984, Sediment trap studies in Lake Michigan: Resuspension and chemical fluxes in the southern basin: Journal of Great Lakes Research,

343 v. 10, p. 307–321, accessed January 10, 2015, at http://www.glerl.noaa.gov/pubs/fulltext/1984/19840004.pdf. Efroymson, R.A., Nicolette, J.P., Suter, II, G.W., 2003, A framework for Net Environmental Benefit Analysis for remediation or restoration of petroleum-contaminated sites: Oak Ridge National Laboratory ORNL/TM-2003/17. Enbridge Energy, L.P., 2010, Work plan for permanent recovery of submerged oil and oil-contaminated sediments at priority locations and Ceresco Dam dredging: Enbridge Line 6B MP 608, Enbridge Energy, L.P., Marshall, Mi., 21 p. (Submitted October 7, 2010). Enbridge Energy, L.P., 2011a, Addendum to the response plan for downstream impacted areas, August 2, 2010 (Revised August 17, 2010 per U.S. Environmental Protection Agency August 17, 2010 letter), Supplement to source area response plan, and supplement to response plan for downstream impacted areas, referred to as operations and maintenance work plan: Enbridge Line 6B MP 608, Enbridge Energy, L.P., Marshall, Mich., 376 p. (Submitted December 4, 2011). Enbridge Energy, L.P., 2011b, Submerged oil recovery standard operating procedure: Enbridge Line 6B MP 608, Enbridge Energy, L.P., Marshall, Mich., Prepared for Michigan Department of Environmental Quality, June 28, 2011. Enbridge Energy, L.P., 2012a, Kalamazoo River hydrodynamic and sediment transport model: Enbridge Line 6B MP 608, Marshall, Mich., 70 p., attachments. (Submitted April 20, 2012) Enbridge Energy, L.P., 2012b, 2012 Morrow Lake Delta and Morrow Lake monitoring and management work plan, Enbridge Line 6B MP 608, Marshall, Mich., November 15, 2012. AR-1720. Enbridge Energy, L.P., 2013a, Sediment poling standard operating procedure: Enbridge Line 6B MP 608, Marshall, Mich., 11 p. (Submitted May 22, 2013). Enbridge Energy, L.P., 2013b, Michigan Department of Environmental Quality Permit 13-39-0012-P, April 30, 2013. Enbridge Energy Partners, L.P., 2013, 2013 crude characteristics, Enbridge Pipelines, Inc., No. 44, accessed January 10, 2015, at http://www.enbridge.com/DeliveringEnergy/Shippers/CrudeOilCharacteristics.aspx. Energy East Pipeline, Ltd., 2014, Oil fate and transport effects, volume 6: Accidents and malfunctions, Section 3: Crude oil characteristics, environmental fate, transport, and effects: Energy East Pipeline Project, 48 p., accessed March 23, 2015 at https://docs.neb-one.gc.ca/ll- eng/llisapi.dll/fetch/2000/90464/90552/2432218/2540913/2543426/2543068/ESA_V6_S3_OilFateTra nspEffects_-_A4E1F4.pdf?nodeid=2543560&vernum=-2. Environment Canada, 2013, Considerations, Volume 2, Report of the Joint Review Panel for the Enbridge Northern Gateway Project, National Energy Board, Calgary, Alberta, 425 p., accessed January 10, 2015, at http://gatewaypanel.review-examen.gc.ca/clf- nsi/dcmnt/rcmndtnsrprt/rcmndtnsrprtvlm2-eng.pdf. Erftemeijer, P.L.A., and Lewis, R.R.R. III, 2006, Environmental impacts of dredging on seagrasses: A review: Marine Pollution Bulletin, v. 52, p. 1553–1572. Fitzpatrick, F.A., Graan, T., Aboulafia, I., Capone, D., Bejarano, A., DeLong, M., Hamilton, S., Michel, J., and Williams, L., 2013, Application and Integration of Net Environmental Benefit Analysis (NEBA) with Tactical Areas: Ceresco Impoundment, Mill Ponds, and Morrow Lake Delta, December 2012 Update in Support of the Kalamazoo River NEBA: Marshall, MI, AR-1710.

344 Fingas, M.F., and Hollebone, B.P., 2003, Review of behavior of oil in freezing environments: Marine Pollution Bulletin v. 47, p. 333–340. Frittelli, J., Andrews, A., Parfomak, P.W., Pirog, R., Ramseur, J.L., Ratner, M., 2014, U.S. rail transportation of crude oil: background and issues for Congress: Congressional Research Service, December 4, 2014, accessed January 10, 2015, at http://fas.org/sgp/crs/misc/R43390.pdf. Geng, X., Boufadel, M.C., and Wrenn, B., 2013, Mathematical modeling of the biodegradation of residual hydrocarbon in a variably-saturated sand column: Biodegradation, v. 24, no. 2, p. 153–163. Gong, Y., Zhao, X., Cai, Z., O’Reilly, S.E., Hao, X., and Zhao, D., 2013, A review of oil, dispersed oil and sediment interactions in the aquatic environment: influence on the fate, transport, and remediation of oil spills: Marine Pollution Bulletin, v. 79, no. 1–2, p. 16–33. Gouvernement du Quebec, Depot Legal, 2014, Tragedie ferroviaire de Lac-Mégantic, Rapport du comite expert sur la contamination residuelle, de la riviere Chaudière par les hydrocarbures petroliers: Gouvernement du Quebec, Quebec, 62 p. ISBN: 978-2-550-70519-2. Grant, A.C., Levy, E.M., Lee, K., and Moffatt, J.D., 1986, PISCES IV research submersible finds oil on Baffin shelf: Current Research, Part A, Part A: Geological Survey Canada, 86-1A, p. 65–69. Great Lakes Environmental Center, Inc., 2012, Preliminary 10-day whole sediment toxicity test report Hyalella azteca and Chironomus dilutus Kalamazoo River sediment sampling, February 2012 Line 6B Spill, Marshall, Michigan: Traverse City, Mich., Great Lakes Environmental Center, March 21, 2012, 8 p. + appendixes. Hamrick, J. M., 1992, A three-dimensional environmental fluid dynamics computer code: theoretical and computational aspects: Gloucester Point, Va., Virginia institute of Marine Science, Applied Marine Science and Ocean Engineering, Special Report No. 317, p. 1–64. Hansen, K.A., 2014, Submerged oil response: Proceedings of the Marine Safety & Security Council, U.S. Coast Guard Journal of Safety & Security at Sea, Winter 2013–2014, p. 22–25. Hoard, C.J., Fowler, K.K., Kim, M.H., Menke, C.D., Morlock, S.E., Peppler, M.C., Rachol, C.M., and Whitehead, M.T., 2010, Flood-inundation maps for a 15-mile reach of the Kalamazoo River from Marshall to Battle Creek, Michigan, 2010: U.S. Geological Survey Scientific Investigations Map 3135, 6 page pamphlet, 6 sheets, scale 1:100,000. James, S.C., Jones, C., and Roberts, J.D., 2005, Consequence management, recovery & restoration after a contamination event: Sandia National Laboratory Report SAND2005-6797. Jones, C., and Lick, W., 2001, SEDZLJ, A sediment transport model: Final Report, Santa Barbara, Calif.: Department of Mechanical and Environmental Engineering, University of California, May 29, 2001. Kaku, V.J., Boufadel, M.C., Venosa, A.D., 2006, Evaluation of mixing energy in laboratory flasks used for dispersant effectiveness testing: Journal of Environmental Engineering, v. 132, no. 1, p. 93–101. Kaplan, E.H., Welker, J.R., and Kraus, M.G., 1974, Some effects of dredging on populations of macrobenthic organisms: Fishery Bulletin, v. 72, no. 2, p. 445–480. Khelifa, A., Stoffyn-Egli, P., Hill, P.S., Lee, K., 2002, Characteristics of oil droplets stabilized by mineral particles: effect of oil types and temperature: Spill Science and Technology Bulletin, v. 8, no. 1, p. 19–30. Khelifa, A., Hill, P.S., Lee, K, 2005a, The role of oil-sediment aggregation in dispersion and biodegradation of spilled oil, in Al-Azab, M., El-Shorbagy, W., and Al-Ghais, S., eds., Oil Pollution and its Environmental Impact in the Arabian Gulf Region: Chapter 10, p. 131–145.

345 Khelifa, A., Hill, P.S., Lee, K., 2005b, A comprehensive numerical approach to predict oil-mineral aggregate (OMA) formation following oil spills in aquatic environments: Proceedings International Oil Spill Conference, May 2005, http://ioscproceedings.org/doi/abs/10.7901/2169-3358-2005-1-873. Khelifa, A., Ajijolaiya, L.O., MacPherson, P., Lee, K., Hill, P.S., Gharbi, S., and Blouin, M., 2005c, Validation of OMA formation in cold brackish and sea waters: Proceedings of the 28th Arctic and Marine Oilspill Program (AMOP) Technical Seminar on Environmental Contamination and Response, Environment Canada, Ottawa, Ontario, p. 527–538. Kim, M., Yim, U.H., Hong, S.H., Jung, J.H., Choi, H.W., An, J., Won, J., and Shim, W.J., 2010, Hebei Spirit oil spill monitored on site by fluorometric detection of residual oil in coastal waters off Taean, Korea: Marine Pollution Bulletin, v. 60, no. 3, p. 383–389. King, T.L., Robinson, B., Boufadel, M., and Lee, K., 2014, Flume tank studies to elucidate the fate and behavior of diluted bitumen spilled at sea: Marine Pollution Bulletin, v. 83, p. 32–37. King, T., Robinson, B., C. McIntyre, S. Ryan, P. Toole, F. Saleh, M. C. Boufadel, and K. Lee, 2015, Fate of surface spills of Cold Lake Blend diluted bitumen treated with dispersant and minerals fines in a wave tank: Environmental Engineering Science. Lee, K., 2000, In situ bioremediation of oiled shoreline environments: Opportunities for Advancement of Environmental Applications of Marine Biotechnology, Proceedings of the October 5-6, 1999, National Research Council of the National Academy of Sciences and the National Academy of Engineering, Washington, DC., National Academy Press, p. 44–60. Lee, K., 2002, Oil-particle interactions in aquatic environments: influence on the transport, fate, effect, and remediation of oil spills: Spill Science and Technology Bulletin, v. 8, no. 1, p. 3–8. Lee, K., Boudreau, M., Bugden, J., Burridge, L., Cobanli, S.E., Courtenay, S., Grenon, S., Hollebone, B., Kepkay, P. Li, Z., Lyons, M. Niu, H. King, T.L., MacDonald, S., McIntyre, E.C., Robinson, B., Ryan S.A., and Wohlgeschaffen, G., 2011, State of knowledge review of fate and effect of oil in the arctic marine environment: Report for the National Energy Board of Canada, Arctic Roundtable: State-of-knowledge review of fate and effects of oil in arctic offshore: Fisheries and Oceans Canada, 259 p., accessed on January 14, 2015, at https://docs.neb-one.gc.ca/ll- eng/llisapi.dll/fetch/2000/90463/621169/700096/704342/A2A8R2_- _NEB_State_of_Knowledge_Review_of_Fate_and_Effect_of_Oil_DFO_COOGER.pdf?nodeid=7043 43&vernum=0. Lee, K., Bugden, J., Cobanli, S., King, T., McIntyre, C., Robinson, B., Ryan, S., and Wohlgeschaffen, G., 2012, UV-epifluorescence microscopy analysis of sediments recovered from the Kalamazoo River: Dartmouth, Nova Scotia, Centre for Offshore Oil, Gas and Energy Research (COOGER) Report, October 24, 2012. Lee, K., and Jarvis, R., 2004, Scientific journal dedicates issue to COOGER research: The COOGER update, March 2004, v. 1, no. 1, p. 3. Lee, K., Li, Z., Kepkay, P., Boufadel, M.C., and Venosa, A.D., 2008, Effects of chemical dispersants on oil-mineral-aggregation in a wave tank: Proceedings of the 2008 International Oil Spill Conference, Savannah, Georgia, p. 633–638. Lee, K., Li, Z., Niu, H., Kepkay, P., Zheng, Y., Boufadel, M., and Chen, Z., 2009, Enhancement of oil- mineral-aggregate formation to mitigate oil spills in offshore oil and gas activities: Final Report (Contract No. M07PC13035) submitted to Minerals Management Service, March 30, 2009, 91 p. Lee, K., Li, Z., Robinson, B., Kepkay, P.E., Blouin, M., and Doyon, B., 2011b, Oil spill countermeasures in the Arctic: Proceedings of the International Conference on Oil Spill Risk

346 Management: Preparedness, Response and Contingency Planning in the Shipping and Offshore Industries, 7–9 March, Malmo Borshus, Sweden, Bellefontaine, Neil, and Linden, Olof, eds., WMU Publications, p. 93–108. Lee, K, Li, Z., Robinson, B., Kepkay, P.E., Blouin, M., and Doyon, B., 2011b, Field trials of in-situ oil spill countermeasures in ice-infested waters: Proceedings of the International Oil Spill Conference, p. 1-16, accessed on January 14, 2015, at http://ioscproceedings.org/doi/pdf/10.7901/2169-3358-2011- 1-160. Lee, K., Lunel, T., Wood, P., Swannell, R., and Stoffyn-Egli, P., 1997, Shoreline clean up by acceleration of clay-oil flocculation processes: Proceedings of the 1997 International Oil Spill Conference (Prevention, Behavior, Control and Cleanup), Fort Lauderdale, Florida, p. 235–240. Lee, K., and Merlin, F.X., 1999, Bioremediation of oil on shoreline environments: development of techniques and guidelines: Pure Applied Chemistry, v. 71, no. 1, p. 161–171. Lee, K., and Stoffyn-Egli, P., 2001, Characterization of oil-mineral aggregates: Proceedings of the 2001 International Oil Spill Conference, Tampa Florida, p. 991–996. Lee, K., Stoffyn-Egli, P., and Owens, E. H., 2002, The OSSA II pipeline oil spill: Natural mitigation of a riverine oil spill by oil-mineral aggregate formation: Spill Science & Technology Bulletin, v. 7, p. 149–154. Lee, K., Stoffyn-Egli, P., and Owens, E.H., 2001, Natural dispersion of oil in a freshwater ecosystem: Desaguadero pipeline spill, Bolivia: Proceedings of the 2001 International Oil Spill Conference, Tampa Florida, p. 1445–1448. Lee, K., Stoffyn-Egli, P., Tremblay, G.H., Owens, E.H., Sergy, G.A., Guénette, C.C., and Prince, R.C., 2003a, Oil-mineral aggregate formation on oiled beaches: Natural attenuation and sediment relocation: Spill Science and Technology Bulletin, v. 8, p. 285–296. Lee, K., Stoffyn-Egli, P., Wood, P.A., Lunel, T., 1998, Formation and structure of oil-mineral fines aggregates in coastal environments: Proceedings of the 21st Arctic and Marine Oilspill Program (AMOP) Technical Seminar, Environment Canada, Ottawa, Ontario, p. 911–921. Lee, K., Weise, A.M., and St-Pierre, S., 1996, Enhanced oil biodegradation with mineral fine interaction: Spill Science and Technology Bulletin, v. 3, no. 4, p. 263–267. Lee, K., Wohlgeschaffen, G., Tremblay, G.H., Johnson, B.T., Sergy, G.A., Prince, R.C., Guénette, C.C., and Owens, E.H., 2003b, Toxicity evaluation with the Microtox test to assess the impact of in-situ oiled shoreline treatment options: Natural attenuation and sediment relocation: Spill Science and Technology Bulletin, v. 8, p. 273–284. Lee, K., Wong, C.S., Cretney, W.J., Whitney, F.A., Parsons, T.R., Lalli, C.M., and Wu, J., 1985, Microbial response to crude oil and Corexit 9527: SEAFLUXES enclosure study: Microbial Ecology, v. 11, p. 337–351. Li, Z., Kepkay, P., Lee, K., King, T., Boufadel, M., and Venosa, A., 2007, Effects of chemical dispersants and mineral fines on crude oil dispersion in a wave tank under breaking waves: Marine Pollution Bulletin, v. 54, no. 7, p. 983–993. Long, E.R., Field, L.J., and MacDonald D.D., 1998, Predicting toxicity in marine sediments with numerical sediment quality guidelines: Environmental Toxicology and Chemistry, v. 17, p. 714–727. Ma, X., Cogswell, A., Li, Z., Lee, K., 2008, Particle size analysis of dispersed oil and oil-mineral aggregates with an automated epifluorescence microscopy system: Environmental Technology, v. 29, no. 7, p. 739–748.

347 MacDonald, D.D., Ingersoll, C.G., and Berger, T.A., 2000, Development and evaluation of consensus- based sediment quality guidelines for freshwater ecosystems: Archives of Environmental Contamination and Toxicology, v. 39, p. 20–31. McFarlin, K.M., Perkins, R.A., Gardiner, W.W., Word, J.D., and Word, J.Q., 2011, Toxicity of physically and chemically dispersed oil to selected arctic species: Proceedings of the 2011 International Oil Spill Conference, Portland, OR, 7 p. McFarlin, K.M., Leigh, M.B., and Perkins, R.A., 2014, Biodegradation of oil and dispersed oil by Arctic marine microorganisms: Proceedings of the 2014 International Oil Spill Conference, Savannah, GA., Poster. McLinn, E.L., and Stolzenburg, T.R., 2009, Ebullition-facilitated transport of manufactured gas plant tar from contaminated sediment: Environmental Toxicology and Chemistry, v. 23, no. 11, p. 2298–2306. Morton, J.W., 1977, Ecological effects of dredging and dredge spoil disposal: A literature review: U.S. Fish and Wildlife Service, Technical Papers of the U.S. Fish and Wildlife Service 94, 33 p. Murdoch, A., and MacKnight, S.D., 1994, Handbook of techniques for aquatic sediments sampling (2d. ed.): Boca Raton, Fla., CRC Press, 256 p. Muschenheim, D.K., and Lee, K., 2002, Removal of oil from the sea surface through particulate interactions: review and prospectus: Spill Science and Technology Bulletin, v. 8, p. 9–18. National Research Council, National Academy of Sciences, 1999, Spills of nonfloating oils: Risk and response: Washington, D.C., National Academy of Science, National Research Council, 87 p. National Research Council, National Academy of Sciences, 2005, Oil spill dispersants, efficacy and effects: Washington, D.C., National Academies Press, 396 p. National Research Council, National Academy of Sciences, 2013, An ecosystem services approach to assessing the impacts of the Deepwater Horizon oil spill in the Gulf of Mexico: Washington, D.C., National Academy of Science, National Research Council, 350 p. Niu, H., Lee, K., Boufadel, M.C., Zhao, L., and Robinson, B., 2014, In situ oil spill countermeasures in ice-infested waters: A modeling study of the fate/behaviors of spilled oil: Proceedings 2014 International Oil Spill Conference, Savannah, GA., p. 1215–1226. Niu, H., Li, Z., Lee, K., Kepkay, P., and Mullin, J.V., 2011, Modelling the transport of oil-mineral- aggregates (OMAs) in the marine environment and assessment of their potential risks: Environmental Modeling and Assessment, v. 16, p. 61–75. Niu, H., Li, Z., Lee, K., Kepkay, P., and Mullin, J.M., 2010, A method for assessing environmental risks of oil-mineral-aggregates to benthic organisms: Human and Ecological Risk Assessment, v. 16, no. 4, p. 762–782. Omotoso, O.E., Munoz, V.A., and Mikula, R.J., 2002, Mechanisms of crude oil–mineral interactions: Spill Science and Technology Bulletin, v. 8, no. 1, p. 45–54. OSPAR, 2008, Literature review on the impacts of dredged sediment disposal at sea: OSPAR Commission, Biodiversity Series, accessed January 15, 2015, at http://www.ospar.org. Owens, E.H., and Lee, K., 2003, Interaction of oil and mineral fines on shorelines: review and assessment: Marine Pollution Bulletin, v. 47, p. 397–405. Passow, U., Ziervogel, K., Asper, V., and Diercks, A., 2012, Marine snow formation in the aftermath of the Deepwater Horizon oil spill in the Gulf of Mexico: Environmental Research Letters, v. 7, p. 1–11.

348 Payne, J.R., Clayton J.R., Jr., Kirstein, B.E., 2003. Oil/suspended particulate material interactions and sedimentation Spill Science Technology Bulletin, v. 8, p. 201–221. Perez, S., Furlan, P., Hussein, N., Shinn, D., and Crook, R., 2014, Interaction between oil and suspended sediments in class 1-2 rivers: Proceedings, International Oil Spill Conference, Savannah, GA, May 2014, v. 2014, no. 1, p. 299120. Perkey, D.W., Smith, S.J., and Kirklin, T., 2014, Cohesive sediment erosion field study: Kalamazoo River, Kalamazoo, Michigan: Vicksburg, Miss., U.S. Army Corps of Engineers Research and Development Center, Coastal and Hydraulics Laboratory, Letter Report, 16 p. + appendixes. Phillips, J.M., Russell, M.A., and Walling, D.E., 2000, Time-integrated sampling of fluvial suspended sediment: A simple methodology for small catchments: Hydrological Processes, v. 14, p. 2589–2602. Pister, B., Lewis, R.D., and Wiese, K., 2009, Site specific geographic response plan: Cabrillo National Monument, Point Loma, California: Cabrillo National Monument Report, 27 p. Prince, M.J., and Blanch, H.W., 1990, Bubble coalescence and break‐up in air‐sparged bubble columns: American Institute of Chemical Engineers Journal, v. 36, no. 10, p. 1485–1499. Rayburn, T., Whelan, A., Jaster, M., Wingrove, R., 2004, Net Environmental Benefit Analysis for Isle Royale National Park, Final Report: Proceedings from the workshop held January 6-8, 2004, Duluth, MN. Reddy, C.M., Arey, J.S., Seewald, J.S., Sylva, S.P., Lemkau, K.L., Nelson, R.K., Carmichael, C.A., McIntyre, C.P., Fenwick, J., Ventura, G.T., Van Mooy, B.A.S., and Camilli, R., 2012, Composition and fate of gas and oil released to the water column during the Deepwater Horizon oil spill: Proceedings of the National Academy of Sciences of the United States of America, v. 109, no. 50. Rymell, M., 2009, RP595, Sunken and submerged oils—behavior and response, A report for the Maritime and Coastguard Agency: BMT Cordah Limited, February, 2009. Sia Partners Energy Outlook, 2011, Oil sands: the oil of the post-oil?: Sia Partners Energy Outlook Weblog, 8-24-2010, accessed January 15, 2015, at http://energy.sia-partners.com/20100824/oil-sands- the-post-oil-oil/. Silliman, B., 2014, Guidelines to prepare for oil sands product spills in varied aquatic environments: Proceedings, 2014 International Oil Spill Conference, Savannah, GA, May 7–9, 2014, p. 426–433. Sterling, M.C., Jr., Bonner, J.S., Ernest, A.N.S., Page, C.A., and Autenrieth, R.L., 2005, Application of fractal flocculation and vertical transport model to aquatic sol–sediment systems: Water Resources, v. 39, p. 1818–1830. Stoffyn-Egli, P., and Lee, K., 2002, Characterization of oil-mineral aggregates: Spill Science and Technology Bulletin, v.8, no. 1, p. 31–44. Sun, J., and Zheng, X.L., 2009, A review of oil-suspended particulate matter aggregation natural process of cleansing spilled oil in the aquatic environment: Journal of Environmental Monitoring, v. 11, p. 1801–1809. Thanh, P.H.X., Grace, M.D., and James, S.C., 2008, Sandia National Laboratories Environmental Fluid Dynamics Code: Sediment Transport User Manual: Sandia National Laboratory Report SAND2008- 5621. Thomas, J.D., Lutz, M.A., Bruce, J.L., Graczyk, D.J., Richards, K.D., Krabbenhoft, D.P., Westenbroek, S.M., Scudder, B.C., Sullivan, D.J., and Bell, A.H., 2007, Water-quality characteristics for selected sites within the Milwaukee Metropolitan Sewerage District Planning Area, Wisconsin, February 2004-September 2005: U.S. Geological Survey Scientific Investigations Report 2007–5084, 187 p.

349 Tsouris, C., and Tavlarides, L., 1994, Breakage and coalescence models for drops in turbulent dispersions: American Institute of Chemical Engineers Journal, v. 40, no. 3, p. 395–406. U.S. Environmental Protection Agency, 1993, Use of chemical dispersants for marine oil spills: Edison, New Jersey, U.S. Environmental Protection Agency Office of Research and Development, EPA/600/R-93/195. U.S. Environmental Protection Agency, 2003, Procedures for the derivation of equilibrium partitioning sediment benchmarks (ESBs) for the protection of benthic organisms: PAH Mixtures: Washington, D.C., U.S. Environmental Protection Agency Office of Research and Development, EPA-600- R-02- 013. U.S. Environmental Protection Agency, 2005, Contaminated sediment remediation guidance for hazardous waste sites: U.S. Environmental Protection Agency Office of Solid Waste and Emergency Response, EPA-540-R-05-012 [variable paginated]. U.S. Environmental Protection Agency, 2011, Cleanup continues; focus on submerged oil: U.S. Environmental Protection Agency Fact Sheet, July 2010 Enbridge Oil Spill, August 2011, 2 p., accessed January 15, 2015, at http://www.epa.gov/enbridgespill/pdfs/enbridge_fs_20110811.pdf. Venosa, A.D., Anasta, P.T., Barron, M.G., Conmy, R.N., Greenberg, M.S., and Wilson, G.J., 2014, Science-based decision making on the use of dispersants in the Deepwater Horizon oil spill, Chapter 1, in Somasundaran, P., Patra, P., Farinato, R.S., Papadopoulos, K., eds., Oil spill remediation: Colloid chemistry-based principles and solutions, (1st ed.): New York, John Wiley and Sons, Inc., p. 1–17. Wang, C., and Calabrese, R.V., 1986, Drop breakup in turbulent stirred‐tank contactors, Part II: Relative influence of viscosity and interfacial tension: American Institute of Chemical Engineers Journal, v. 32, no. 4, p. 667–676. Wang, W., Zheng, Y., and Lee, K., 2013, Chemical dispersion of oil with mineral fines in a low temperature environment: Marine Pollution Bulletin, v. 72, p. 205–212. Wang, W., Zheng, Y., Li, Z., and Lee, K., 2011, PIV investigation of oil–mineral interaction for an oil spill application: Chemical Engineering Journal, v. 170, p. 241–249. Waters, T.F., 1995, Sediment in streams: sources, biological effects, and control: Bethesda, Maryland, American Fisheries Society Monograph 7, 251 p. Weaver, J.W., 2004, Characteristics of spilled oils, fuels, and petroleum products, 3a, Simulation of oil spills and dispersants under conditions of uncertainty: Washington, D.C., U.S. Environmental Protection Agency, EPA 600/R-04/120. Weise, A.M., Nalewajko, C., and Lee, K., 1999, Oil-mineral fine interactions facilitate oil biodegradation in seawater: Environmental Technology, v. 20, p. 811–824. Wood, P., Lunel, T., Bailey, N., Lee, K., and Stoffyn-Egli, P., 1997, Clay-oil flocculation during surf washing at the Sea Empress incident: Proceedings of the 20th Arctic and Marine Oil Spill Program Technical Seminar, p. 1,085–1,105. Zhang, H., Khatibi, M., Zheng, Y., Lee, K., Zhengkai, L., Mullin, J., 2010, Investigation of OMA formation and the effect of minerals: Marine Pollution Bulletin, v. 60, p. 1433–1441. Zhao, L., Torlapati, J., Boufadel, M., King, T., Robinson, B., and Lee, K., 2014a, VDROP: a numerical model for the simulation of droplet formation from oils of various viscosities: Chemical Engineering Journal, v. 253, p. 93–106.

350 Zhao, L., Torlapati, J., King, B. Davidson, T., Boufadel, M., and Lee, K., 2014b, A numerical model to simulate the droplet formation process resulting from the release of diluted bitumen products in marine environment: Proceedings, 2014 International Oil Spill Conference, Savannah, GA, p. 449– 462.

351 Fitzpatrick and others—Oil-Particle Interactions and Submergence from Crude Oil Spills in Marine and Freshwater Environments— Review of the Science and Future Science Needs—Open-File Report 2015−1076 352 MARINE ECOLOGY PROGRESS SERIES Vol. 316: 185-199, 2006 Mar Ecol Prog Ser Published July 3

Selective foraging by fish-eating killer whales Orcinus orca in British Columbia

John K. B. Ford*, Graeme M. Ellis

Fisheries and Oceans Canada, Pacific Biological Station, 3190 Hammond Bay Road, Nanaimo, British Columbia V9T 6N7, Canada

ABSTRACT: As the apex non-human marine predator, the killer whale 0rdnus orca feeds on a wide diversity of marine fauna. Different ecotypic forms of the species, which often exist in sympatry, may have distinct foraging specialisations. One form found in coastal waters of the temperate NE Pacific Ocean, known as the 'resident' ecotype, feeds predominantly on salmonid prey. An earlier study that used opportunistic collection of prey remains from kill sites as an indicator of predation rates suggested that resident killer whales may forage selectively for chinook salmon 0ncorhynchus tshawytscha, the largest but one of the least abundant Pacific salmon species. Potential biases in the prey fragment sampling technique, however, made the validity of this finding uncertain. We under took field studies of foraging behaviour of resident killer whales to resolve this uncertainty and to examine potential variation in prey selection by season, geographical area, group membership and prey availability. Foraging by resident killer whales was found to frequently involve sharing by 2 or more whales. Prey fragments left at kill sites resulted mostly from handling and breaking up of prey for sharing, and all species and sizes of salmonids were shared. Resident killer whale groups in all parts of the study area foraged selectively for chinook salmon, probably because of the species' large size, high lipid content, and year-round availability in the whales' range. Chum salmon 0ncorhynchus keta, the second largest salmonid, were also taken when available, but smaller sock eye 0. nerka and pink 0. gorbuscha salmon were not significant prey despite far greater seasonal abundance. Strong selectivity for chinook salmon by resident killer whales probably has a significant influence on foraging tactics and seasonal movements, and also may have important implications for the conservation and management of both predator and prey.

KEY WORDS: Prey choice · Food sharing · Foraging specialisation · Salmonid predation ------Resale orrepublication not permitted without writtenconsent of the publisher------

INTRODUCTION caused the depletion of several marine mammal popu lations in the North Pacific in the post-whaling era Marine mammals that occupy high trophic positions (Springer et al. 2003). A variety of arguments have been in marine ecosystems can have important top-down raised against this hypothesis (e.g. DeMaster et al. effects on those ecosystems (Bowen 1997). As the apex 2006, Mizroch & Rice 2006), which emphasise how (non-human) marine predator, the killer whale 0rdnus much uncertainty exists regarding the dynamics of orca is capable of preying on a great variety of species killer whales and their prey. Developing an under and has a remarkably diverse diet. The list of prey standing of the factors involved in prey selection, as taken by this cosmopolitan predator includes more than well as the extent of foraging specialisation or flexibil 120 species of fishes, cephalopods, sea turtles, sea ity, will be needed if we are to better understand the birds, mustelids, pinnipeds and cetaceans (Martinez & roles played by killer whales in marine ecosystems. Klinghammer 1970, Jefferson et al. 1991, Matkin & Such knowledge is also important for conservation of Saulitis 1994, Fertl et al. 1996, Simila et al. 1996, Ford et both killer whale populations and their prey. al. 1998). With such a diversity of prey types, killer Although a generalist as a species, different killer whales can potentially influence ecosystem structure whale populations can have strikingly divergent forag and function at a variety of levels. For example, it has ing specialisations. For example, in coastal waters of the recently been proposed that predation by killer whales NE Pacific Ocean, 2 sympatric, genetically-distinct killer

353 186 Mar Ecol Prog Ser 316: 185-199, 2006

whale ecotypes, so-called 'residents' and 'transients', mon salmonids during the summer migration period, feed almost exclusively on fish and marine mammal represented almost two-thirds of prey items identified. prey, respectively (Bigg et al. 1990, Baird & Dill 1995, The disproportionate occurrence of chinook salmon in Ford et al. 1998, 2000, Hoelzel et al. 1998, Ford & Ellis prey samples relative to its low abundance led Ford et al. 1999, Barrett-Lennard 2000). Killer whale populations in (1998) to suggest that resident killer whales may forage other regions, such as Norway, Argentina and Antarc selectively for this species over other available salmo tica, also specialise on particular prey types, and often nids. Chinook is the largest salmonid and has a relatively employ elaborate foraging tactics in order to do so high lipid content, features that may be desirable for for (Lopez & Lopez 1985, Simila & Ugarte 1993, Baird 2000, aging killer whales. However, Ford et al. (1998) also Pitman & Ensor 2003). Such specialisations appear to raised concerns that potential biases in their prey sam represent behavioural traditions that are passed across pling method may have led to over-representation of chi generations by social learning (Guinet & Bouvier 1995, nook and under-representation of other species in sam Ford et al. 1998, Saulitis et al. 2000), and have been ples from predation events. Chief among these was the described as cultures (Rendell & Whitehead 2001). possibility that chinook, being larger than other sal Behavioural traditions may determine the overall monids, may be broken up by whales prior to being type of prey that is acceptable to a particular killer eaten, thus shedding more scales than smaller salmonids whale population and the foraging tactics employed in the process (Ford et al. 1998). However, without for prey capture, but other factors are likely to play knowledge of the details of prey handling and consump important proximate roles in prey selection. Prey tion of salmonid and other fish species by resident killer choice by predators is influenced by rates of encoun whales, it was not possible to evaluate the significance, ters with a prey species and its profitability, which is if any, of this potential bias. It was concluded that resi determined by the prey item's net energy value and dent whales may have a preference for chinook, but the the amount of time needed to catch and handle it extent of their selectivity for this species remained uncer (Stephens & Krebs 1986, Scheel 1993, Bowen et al. tain (Ford et al. 1998). These potential biases as well as 2002). Factors that are important in prey choice in small sample sizes also prevented Ford et al. (1998) from killer whale populations are poorly known, but long examining potential differences in prey selection among term studies in the NE Pacific have provided some different resident communities or social groups, as sug insights. Mammal-hunting transient killer whales in gested by Nichol and Shackleton (1996), or by different this region prey on at least 9 species of marine mam sex or age classes, as suggested by Bain (1989). mals, including seals, sea lions, porpoises, dolphins In order to address these and other gaps in the cur and baleen whales, but appear to prefer small species rent knowledge of resident killer whale diet and prey such as harbour seals Phoca vitulina and harbour por selection, we undertook field studies of foraging poises Phocoena phocoena, which are common year behaviour and feeding by resident killer whales from round in the whales' range (Haley 1986). These species 1997 to 2005 to build upon the data presented in Ford are relatively easy to capture and kill, and have a low et al. (1998). In particular, field efforts from 2003 to probability of causing injury to the attacking whales 2005 were focused on documenting the detailed (Baird & Dill 1995, Ford et al. 1998, 2005). Sympatric, aspects of prey capture and handling to assess the fish-eating resident killer whales show movement pat validity of using prey fragment sampling to interpret terns in nearshore waters that are closely associated dietary preferences. In this report, we present new with high densities of migrating salmon (Heimlich information on the frequent occurrence of cooperative Boran 1986, Guinet 1990, Nichol & Shackleton 1996). foraging and prey sharing in resident killer whales, Nichol & Shackleton (1996) found positive correlations and the implications of this behaviour with respect to between the seasonal occurrence of resident killer the use of prey fragments as indicators of prey selec whales and 3 of the most abundant salmonid species, tion. Extensive sampling of feeding events also pink salmon 0ncorhynchus gorbuscha, sockeye allowed us to evaluate prey selection quantitatively salmon 0. nerka and chum salmon 0. keta, off NE with regard to prey species availability, as well as by Vancouver Island, British Columbia. Ford et al. (1998) geographical location, group membership, and age confirmed that salmonids were the predominant food and sex class of foraging whales. of resident killer whales in these waters by identifying prey species from scales and tissue fragments collected MATERIALS AND METHODS from kill sites. An unexpected finding, however, was that sockeye, pink and chum salmon, despite being correlated with the occurrence of killer whales in this Study area and population. Field studies on the life area, formed the minority of prey samples. Instead, history, social organisation, acoustic behaviour, and chinook salmon 0. tshawytscha, one of the least com- population genetics of killer whales in British Colum-

354 Ford & Ellis: Selective foraging by killer whales 187

bia have been undertaken since 1973 (Bigg 1982, Bigg these data are included in the present study. In 2003 to et al. 1987, 1990, Olesiuk et al. 1990, Ford 1989, Ford et 2005, field studies were dedicated to systematically al. 1998, 2000, Barrett-Lennard 2000). These long-term documenting foraging behaviour and collecting preda studies have relied extensively on the photographic tion data, in addition to conducting the annual census identification of individuals from natural markings of individuals by photo-identification (Bigg et al. 1987, (Bigg et al. 1987, Ford et al. 2000). In these waters, 2 Ford et al. 2000). A total of 152 field days were devoted communities of resident killer whales, 'northern resi to these objectives in 2003 to 2005. dents' and 'southern residents' can be found in all Field studies from 1974 to 2002 were conducted months of the year, but are observed mostly during using a variety of vessels from 5 to 20 m in length. In May to November. The northern resident community is 2003 to 2005, dedicated focal-individual and focal found typically from mid Vancouver Island to SE group studies were undertaken mainly from a 10 m Alaska, and the southern resident community off the long power vessel. When whales were encountered, southern half of Vancouver Island and in the inland individuals were observed visually or photographed to waters of Washington State. Whales from the 2 com determine the identity of matrilines present. Photo munities have not been seen to associate, despite graphic identification procedures are described in extensive overlap in their ranges (Ford et al. 2000). Bigg et al. (1987) and Ford et al. (2000). Once the iden Large aggregations of resident killer whales can be tity of killer whales present in the encounter was found in certain coastal locations during summer. The established, effort was directed to documenting forag whales greatly reduce their use of these locations in ing behaviour and collecting scales and tissue frag winter and spring, and their range during this period is ments from prey killed during feeding events. The poorly known (Ford et al. 2000, Wiles 2004). The north activity state of the whales was determined from sur ern and southern resident communities contained 219 facing and dispersion patterns (see Ford 1989 for defi and 87 individuals, respectively, in 2004 (authors' nitions of activity states). When foraging, whale groups unpubl. data, K. Balcomb, Center for Whale Research, typically spread out over several square kilometres, Friday Harbor, Washington, unpubl. data). with individuals and subgroups swimming and diving The basic social unit of resident killer whales is the independently but travelling generally in the same 'matriline', which consists of individuals that are closely direction. Surfacing whales were observed by eye or related by matrilineal descent. Matrilines generally binoculars for signs of prey pursuit or capture. When contain an old female, or matriarch, and 1 to 3 genera apparent feeding was observed, the site of the kill was tions of her descendents of both sexes. Dispersal of indi approached quickly, while taking care to avoid dis viduals from the matriline is extremely rare (Ford et al. turbing the whales, in order to determine identity of 2000). Matrilines are comprised of an average of 6 the individual(s) involved and to search for prey frag members (±0.59 SE, range 1 to 26, n = 50). Resident ments in the water. Whether or not feeding was con killer whales typically travel in 'pods', which consist of firmed, the individual or subgroup was then followed related matrilines that spend the majority of their time at distances of 50 to 150 m to document subsequent together (Bigg et al. 1990). Although some pods origi feeding events. Focal individuals and subgroups (Alt nally described in the 1970s and 1980s have maintained mann 1974, Mann 1999) that were actively feeding their stability, others have split in recent years (Ford et were followed for as long as the activity continued or al. 2000). 'Clans' are comprised of pods and matrilines until focal animals joined other groups and could no that have descended from a common matrilineal ances longer be followed individually. tor and have a unique set of shared dialects. The north The behaviour of focal individuals and subgroups ern resident community consists of 3 clans, A, G, and R, was monitored closely and constantly during feeding while the southern resident community is made up of a sessions. Particular attention was given to direction of single clan, J. Members of the northern resident clans travel, regularity of dive durations, and extent of sub frequently associate with one another. group cohesion, as changes in these variables often Field effort and procedures. Data on predation by signalled a feeding event. Individuals or subgroups resident killer whales in British Columbia have been suspected to have captured a prey item were ap collected each year since 1974. Data collected from proached to within 20 m to observe prey handling and 1974 to 2002 consisted mostly of opportunistic observa consumption. To collect evidence of feeding, the sur tions of feeding events and collection of prey frag facing locations of the feeding whale or subgroup were ments from the vicinity of kills. Effort varied widely also examined for prey fragments at the surface or in according to changing research objectives, but preda the water column. The principal observer, who was tion studies were given higher priority after 1990 (Ford also the boat operator, was situated approximately 4 m et al. 1998, 2000). The results of these studies up to above the water surface on the flying bridge of the re 1996 were reported by Ford et al. (1998), and some of search vessel. This position affordeda high-angle view

355 188 Mar Ecol ProgSer 316: 185-199, 2006

into the water as the boat was manoeuvred. A second Table 1. Orcinus area. Evidence for 529 kills by resident killer observer stood on the vessel's bow, holding a fine whales documented from 19?4 to 2005 mesh dip net (mesh size approximately 1 mm) with 4 m % telescoping handle, and also searched for fragments. Evidence of predation No. of kills When fish scales or pieces of tissue were seen, the boat Observation only 80 15.1 was immediately stopped and the net was deployed to Both tissue and scale samples 100 18.9 retrieve the fragments. Fragments were collected Tissue samplesonly 21 4.0 mostly at depths of Oto 2 m, but occasionally as deep as Scale samples only 328 62.0 3 to 4 m in calm conditions with good water clarity. Total 529 100 Generally, only a subsample of the many scales and tis sue fragments seen in the water were collected from each kill. Rain, winds greater than 10 knots, and high of-fit tests. The standard error (SE) of the mean is given water turbidity reduced the success rate of fragment as a measure of variability. location and collection. Prey fragments and scales were stored in 10 ml vials containing 95 % ethanol. Un til 2004, no effort was made to collect tissue samples RESULTS from fish kills when scales were available for collec tion, but both were collected systematically in that year A total of 529 feeding events were observed during and in 2005. The date, time, and geographical position 206 encounters with resident killer whales between (from a GPS instrument) of the feeding event was 1974 and 2005. Of these, 340 (63 %) were recorded recorded, as well as the identity of the individual mak during dedicated feeding studies in 2003 to 2005. Prey ing the kill and other whales involved in the prey cap samples (tissue or fish scales) were recovered from 449 ture or consumption. If individuals could not be identi feeding events (85 %), while the remaining 15% were fied, their age/sex class was noted whenever possible. documented by observation only (Table 1). A mean of Prey species identification and ageing. Many spe 4.8 scales per feeding event (±0.24 SE, range 1 to 28) cies of fishes are readily identifiable at a distance by an were collected from 428 of the 529 feeding events experienced observer, but salmon species can be diffi (81%). During 2004 to 2005, when tissue as well as cult to distinguish without close examination. Al scales were collected systematically when both were though Ford et al. (1998) included salmonid identifica available, tissue fragments were retrieved from 115 of tions based on field observations, in the current 249 (46%) feeding events. analyses we included only positive species identifica Feeding events were documented from May to tions of salmonids based on scales or tissue samples to December. A total of 463 (87.5 %) feeding events eliminate this potential source of error. Fish scales involved northern residents, and 66 (12.5 %) events were analysed by the Fish Ageing Laboratory at the involved southern residents. All 4 resident clans and Pacific Biological Station (Department of Fisheries and all but 1 of the 19 resident pods in the northern and Oceans, Nanaimo, British Columbia, Canada) to deter southern communities are represented in this data set mine species identity and age. Age was designated (the exception is W1 pod; Ford et al. 2000). Samples using the European method, whereby years in fresh were collected from most regions of the coast, but two water after hatching and years in marine water are thirds came from waters off NE Vancouver Island, an identified and separated by a decimal point (Groot & important core area for northern residents (Fig. 1, Margolis 1991), and age class was assigned according Table 2; see also Ford et al. 2000). Kills made during a to a standard 1 January birth date. Age class was thus total of 274 feeding events could be attributed to calculated by summing the freshwater and marine whales of known age or sex class. The monthly distrib years of the European age and adding 1. ution of these is shown in Table 3. Scales that could not be positively identified to species and tissue samples collected from feeding events were submitted to the Molecular Genetics Laboratory at the Foraging behaviour and prey fragment sampling Pacific Biological Station for species identification using analysis of microsatellite DNA. The methodology of Dedicated studies of foraging behaviour of resident these analyses is described in Withler et al. (2004). killer whales in 2003 to 2005 resulted in sampling 331 General analyses. Statistical analyses were con feeding events on 60 d. Within this period, focal sub ducted using SPSS Version 11.0. The significance of groups or individuals were followed and observed for a differences between 2 group means was tested using total of 61.7 h during 34 feeding sessions, which we Mann-Whitney U-tests, and among different frequen defined as the interval between the first and last feed 2 cies of occurrence by chi-squared (x, ) goodness- ing events in a series by that individual or subgroup.

356 Ford & Ellis: Selective foraging by killer whales 189

Table 2. Orcinus area. Species composition of fishes killed in 529 whale feeding events from 1974 to 2005 in different coastal regions of British Columbia (for details see Fig. 1). Species identity of salmonids was determined by scale analysis (n = 412) or from DNA analysis of tissue fragments (n = 20). Species identity of non-salmonid fishes was determined from field observation (n = 1), scale analysis (n = 2), examination of partial carcasses recovered from kills (n = 3), and from DNA analysis of tissue fragments (n = 1). PFMA: Pacific Fisheries Management Areas of Fisheries & Oceans Canada; UnSa: salmonids observed as prey in the field but not sampled, or salmonids that could not be identified to species. UnFi: fishes that could not be positively identified to species and could include either salmonids or non-salmonids. QCI: Queen Charlotte Islands. All samples from PFMA Areas 1 to 13 were collected from northern resident killer whales. Samples from PFMA Areas 14 to 29 involved southern resident killer whales, except for the sablefish sample, which was collected from a northern resident killer whale. Of the total 529 feeding events, 135 were included in Ford et al. (1998)

Region PFMA n 0. tshawytseha 0. keta 0. kisuteh 0. gorbuseha 0. nerka 0. mykiss Other UnSa UnFi (Chinook) (Chum) (Coho) (Pink) (Sockeye) (Steelhead)

North coast and QCI 1-6 72 43 17 0 0 0 0 1a 10 Central coast 7-11 66 47 3 0 0 0 11 3 Vancouver Island NE 12-13 324 177 76 6 12 0 0 2b 44 7 SE 14-19,28-29 47 27 2 2 0 1 2 1C 10 2 w 20-27 20 15 0 0 0 0 0 3d 2 0 Total 529 309 98 9 13 2 7 n 13 % of identified salmonids 71.5 22.7 2.1 3.0 0.2 0.4

"1 Pacific halibut Hippoeampus stenolepis bl yelloweye rockfish Sebastes ruberrimus, 1 herring Clupea pallasi c1 quillback rockfish S. maliger d2 herring C. pallasi, 1 sablefish Anoplopoma fimbria

130 °w Intervals between feeding events ranged ' from 2 to 120 min, with an average of -.. 25.8 min (±1.84 SE, n = 136 intervals). � Almost one-third of feeding events in a � British Columbia session were 10 min or less apart. Based ; iN on the distinct behavioural cues associ- � ated with prey capture and consumption, : tt� a� we believe that few feeding events were • Island (1-6� •• missed during feeding sessions by focal NE Vancouver •c;��;·��-ast Island groups or individuals. (7-11)...... •• (12-13) An average of 4.5 matrilines were pre t sent during encounters in which feeding ° behaviour was documented (± 0.37 SE, 50 N ··· ··· range 1 to 12 matrilines, n = 60 encounters). Subgroups, which usually consisted 0 100 200 SE Vancouver of complete matrilines or partial matri Island km lines comprised of mothers and their (14-19, 28-29) young offspring, often foraged 200 m or more apart. Adult males usually foraged independently or in association with their 130 °w mother, especially in cases where the mother had no juvenile offspring. Whales Fig. 1. Coastal regions of British Columbia (Pacific Fisheries Management often foraged close to shorelines, espe Area designations). Numbers of Orcinusarea feeding events and prey species identified for each region are shown in Table 2 cially in the deep, narrow channels and straits frequented by resident killer whales during summer and autumn. Feeding sessions by focal individuals and subgroups Adult males usually foraged further offshore than sub lasted an average of 1.9 h (± 0.29 SE, range 0.4 to groups. Individuals and subgroups foraging nearshore 9.25 h). Of the total 331 feeding events, 170 were doc followed the coastline closely, often within 50 m of umented during these 34 sessions, with an average of 5 shore. Whales foraging offshore often swam in a zigzag feeding events per session (±0.47 SE, range 2 to 15). pattern rather than in a straight line along a channel.

357 190 Mar Ecol ProgSer 316: 185-199, 2006

Table 3. Orcinus orca. Monthlydistribut ion (for 1974 to 2005) Table 4. Orcinus orca. Frequency of prey sharing in feeding of sampled feeding events where age/sex class of whale events by resident killer whales where age and/or sex class making kill was determined; n = 274 kills of individuals making kills could be determined. Juvenile whales were < 14 yr old Month ---- Age/sex class--- Total Adult male Adult female Juvenile Sharing ---- Age/sex class--- Total Adult male Adult female Juvenile May 8 5 14 Jun 7 2 10 Yes 9 97 36 142 Jul 16 26 25 67 No 44 4 9 57 Aug 19 71 31 121 Total 53 101 45 199 Sep 13 3 6 22 Oct 14 15 11 40 Total n 122 75 274 On numerous occasions, members of a subgroup milled at the surface while one of the group was under water on a long dive. Upon surfacing with prey, the Signs of prey pursuit and capture were at times very whale was seen to carry the fish in the direction of the conspicuous. When an appropriate prey item was milling individuals. Scales and/or tissue were found detected, an individual would break suddenly into a where the whales joined. On other occasions, a whale high-speed chase that continued for 10 to 30 s, or was observed to surface with prey and carry it for 3 to (rarely) up to 3 min. Chases were directional and non 5 shallow dives and surfacings while another whale directional, the latter accompanied by fast turns and swam quickly in its direction. Although small numbers rolls at the surface. Chases often took place along of scales were often seen in the water in the trail of the steep shorelines, with whales swimming at high speed whale carrying the fish, much larger numbers of scales within a few metres of the rocks. Although vigorous and pieces of tissue were observed at the site of join chases were obvious indicators of predation, more ing, indicating that prey consumption was delayed often signs of feeding were quite subtle, and close until the whales were together. attention to several behavioural cues was necessary in Observations during 235 feeding events provided order to detect them. For example, a change in the oth sufficient evidence to judge with reasonable confi erwise consistent swimming speed and direction of for dence whether or not sharing took place. In 57- of these aging whales often indicated that a pursuit was under feeding events (24 %) there was no indication of any way. An unusually long 5 to 7- min dive following an sharing. In the other 17-8 cases (76%), sharing was extended series of regular, 2 to 3 min dives was also a either clearly evident or strongly suspected. Between good indication of prey pursuit and possible capture. 2 and 6 whales (including the individual making the Whale interactions during feeding events indicated kill) were involved in shared feeding events, though it that the majority of prey items were shared by 2 or was generally not possible to determine how many more individuals. Typically, a whale that made a kill individuals actually took part in prey consumption. was joined by others, and the group would swim Whales involved in shared feeding events typically together for 2 to 3 surfacings before splitting up once belonged to the same matriline. The frequency of shar again. Individuals converged on the successful whale ing by members of different age and sex classes 2 from as far as 400 m, although more often joining of whales differed significantly (X = 30.8, df = 2, whales were within 100 to 200 m when the kill took p < 0.001; Table 4), with adult males sharing signifi place. On other occasions, several whales were cantly fewer of their kills (17%) than expected com involved in the pursuit and would work cooperatively pared to adult females (96% ) and juveniles (80 % ) , to take prey that had sought refuge in crevices along (these were not distinguished by sex). rocky shorelines or in kelp beds. After making the kill, members of the group joined at the surface and swam together for several surfacings. Inspection of the site at Prey species and age composition which whales joined in such situations invariably revealed fish scales or pieces of tissue in the water. All 529 feeding events involved fishes, at least 96% Often, as the whales swam together after joining, a of which were salmonids (Table 2). We could not iden trail of prey fragments was left in the water, indicating tify 10 samples (2.5 %) to species, and some of these that the prey item was being torn up along the way. may also have been salmonids. The only non-salmonids Close observations of prey handling and consump identified were 3 Pacific herring Clupea pallasi, 1 tion provided strong evidence that sharing was taking sablefish Anoplopoma fimbria, 1 yelloweye rockfish place in such circumstances, and that intentional pro Sebastes rubenimus, 1 quillback rockfish S. maliger, visioning of other whales was also frequently involved. and 1 Pacific halibut Hippocampus stenolepis.

358 Ford & Ellis: Selective foraging by killer whales 191

Table 5. 0rcinus orca. Ages of 344 salmonids killed by resident killer whales. salmonid kills by adult males was sig Ages given in European system, whereby years in freshwater after hatching and nificantly greater than in kills by adult years in salt water are identified and separated by decimal point. Age classes 2 females and juveniles (x, = 10.4, df = 1, used elsewhere in this paper were obtained by summing the 2 European age values and adding 1 (e.g. 1.2 age converts to a 4th yr fish). Specific names of p < 0.01). This difference may be due to prey in Table 2 a seasonal bias in sampling of kills by different sexes. A greater proportion of Species n ------Age class------kills by males (35 % of total samples 0.1 0.2 0.3 0.4 0.5 1.1 1.2 1.3 1.4 1.5 2.1 2.2 from males; Table 3) than by females (15 % of total, Table 3) were sampled Chinook 236 25 94 49 2 3 16 34 10 during September and October, when Coho 9 8 Chum 87 57 27 3 chum salmon was the predominant Pink 12 12 prey species (Table 7). All 11 pink salmon kills by identified whales were by juveniles. In fact, 6 of the 11 pink Of the 7 species of Pacific salmon (Oncorhynchus salmon kills were made during a single 1.5 h long feed spp.) found in the study area, 6 were represented in ing session by a 1 yr old calf. Because pink salmon are feeding samples (the exception was the cutthroat trout the smallest of the Pacific salmonids (Quinn 2005), we 0. darld). Species identity was determined for 432 examined the age distribution of chinook salmon kills salmonids, 20 by DNA analysis and the remainder by to determine if younger (and thus smaller) salmon scale analysis. The frequency of occurrence of these were taken more frequently by young whales than by species in whale kills is shown by region in Table 2. adults. The mass of chinook salmon increases dramati- Chinook salmon was by far the predominant salmonid cally with increasing age, from a mean of 1.1 kg in 2 yr observed, representing 71.5 % of salmonid kills identi old fish to over 13 kg in 5 to 6 yr old fish (Table 9). fied to species. The second most important salmonid was chum, at 22. 7 % of samples. Coho, pink, sockeye Table 6. 0rcinus orca. Salmonid species sampled from feed and steelhead together represented less than 6 % of ing events by members of different resident clans. Clans A, G the salmonids identified. Ages were determined for and R form the northern community, Clan J is the southern 344 salmonid samples (Table 5). communit y. Sample size (n = 423 kills) differs from Tables 2 & Chinook was the principal species taken by resident 7 because not all kills could be positively attributed to specific individuals or their clans. Specific names of prey in Table 2 killer whales in all regions of the coast (Table 2). It was the most common species in feeding samples from both Species Clan Total northern and southern communities, as well as from A G R J each of the resident clans (Table 6). Chinook re presented the majority of salmonid samples collected Chinook 178 70 19 39 306 from resident killer whales during May to August (Table Chum 64 24 4 2 94 7). However, chum salmon was the predominant species Coho 'J 0 0 2 9 Pink 9 2 0 0 11 identified from feeding events in September to October. Sockeye 0 0 0 1 1 Pink salmon occurred in small numbers in July to Steelhead 0 0 0 2 2 September samples, as did coho salmon during July to Total 258 96 23 46 423 October. Southern residents were responsible for the single sockeye salmon sample, collected in July, and the 2 steelhead samples, Table 7. 0rcinus orca. Salmonid species sampled from resident killer whale collected in November and December. feeding events by month; n = 432 kills. Specific names of prey in Table 2

Month Species Total Prey selection as a function ol'whale Chinook Chum Coho Pink Sockeye Steelhead age/sex class May 22 0 0 0 0 0 22 Jun 34 12 0 0 0 0 46 Adult male, adult female and juve Jul 94 13 1 0 1 0 109 nile (<14 yr old) resident killer whales Aug 139 3 4 11 0 0 157 all preyed on chinook salmon more Sep 11 22 1 2 0 0 36 than on any other species (Table 8). Oct 8 48 3 0 0 0 59 Nov 1 0 0 0 0 2 Chum and coho salmon were also Dec 0 0 0 0 0 taken by each sex and age categ01y, Total 309 98 9 13 2 432 although the proportion of chum in

359 192 Mar Ecol Prog Ser 316: 185-199, 2006

Table 8. Oreinus area. Salmonid species sampled from feed was not significantly different from the mean age of ing events as a function of age and sex class of resident 4.27 yr (± 0.16 SE, n == 22 kills)= of chinook taken in non whales; n = 274 kills. Specific names of prey in Table 2 shared kills (U-test 1260, p 0.87).

Species ----- Age/sex class--- Total Adult male Adult female Juvenile Prey selection versus availability Chinook 4'i' 104 40 191 Chum 26 16 23 65 To assess the extent to which foraging resident killer 1 Coho 4 2 'i' whales select for particular species or sizes of sal Pink 0 0 11 11 monids, the species and age composition of kills sam Total n 122 'i'5 2'i'4 pled off NE Vancouver Island was compared to salmonid availability. The relative abundance of sal Table 9. Onearhynehus tshawytseha. Mean fork lengths and monids was determined from catch statistics resulting mass (±SE) of chinook salmon at different ages. Data from from Fisheries and Oceans Canada seine test fisheries seine caught fish in MarkRecovery Program of Fisheries and Oceans Canada (Kuhn 1988) (Data available from www-ops2.pac.dfo-mpo.gc.ca/ xnet/content/salmon/testfish/default.htm),which were Age (yr) Length (mm) Mean mass (kg) n undertaken concurrently with and in close proximity to our sampling of killer whale feeding events in the 2 425 ±1.19 1.1 ± 0.01 30'i'2 western Johnstone Strait area, off NE Vancouver 3 581 ±2.14 3.1 ± 0.04 3206 Island (see Fig. 1). These data provide a reliable index 4 808 ±3.43 8.5±0.11 91'i' of the relative availability of different salmonid species 5 939± 4.21 13.3 ±0.20 426 6 961 ± 15.0 13., ± o.n 3'i' to the whales foraging in that area during particular periods of the season. Fig. 4 illustrates the species com position of salmonids caught in test fisheries and by Although the smallest chinook (2 and 3 yr olds) were resident killer whales during 3 periods in July to Octo taken more often by juvenile than by adult whales ber 2004. From 15 July to 15 August, the test catches (Fig. 2), the overall trend was not significant ( U-test = 1688, p = 0.11).= There was= also no significant differ 60 ence (U-test 1377.5, p 0.13) between the mean age= □ Adults of chinook taken by adult males (4.50 yr± 0.13 SE, n !:!!. 50 GIJuveniles 38) and by adult females (4.26 yr± 0.08 SE, n = 86). :sz 0 40 (]) Prey species and sizes in shared and non-shared kills 2 30 (]) Of the 6 salmonid species taken by resident killer (]) 20 whales, 4 were identified from both shared and non shared feeding events (Table 10). The great majority 10 (84 %) of chinook salmon tended to be shared, while a significantly lower proportion of chum salmon were 0 2 = = 2 3 4 5 6 shared (55 %, x 4.47, df 1, p < 0.05). Although chi Chinook age (yr) nook are often larger than chum salmon (Healey 1986), Fig. 2. Onearhynehus tshawytseha. Age distribution of chi larger size appears not to be the reason for the greater nook salmon taken by adult (n = 12'i'kills) and juvenile (n = 32 incidence of sharing of chinook. Chum salmon formed kills) resident killer whales Oreinus area a higher proportion of the prey samples from adult males than from adult females, and because males Table 10. Oreinus area. Frequency of prey sharingby resident shared prey less frequently than females, the propor killer whales in feeding events where prey was identified to tion of chum that were not shared may as a result be species. n = 229 kills. Specific names of prey in Table 2 higher than for other species. Both coho and pink salmon, which tend to be smaller than chum (Healey Sharing ----Prey species---- Total 1986), were noted in both shared and non-shared feed Chinook Chum Coho Pink ing events. Also, the frequency of sharing of chinook Yes 141 29 3 1'i'4 salmon did not differ with the age, and hence the size, No 2'i' 24 3 55 of the fish (Fig. 3). The mean age of chinook taken in Total 168 53 6 2 229 shared kills was 4.28 yr (± 0.07 SE, n = 117 kills), which

360 Ford & Ellis: Selective foraging by killer whales 193

were dominated by migrating sockeye and pink 60 □ Shared salmon, which together comprised over 90 % of salmon .!!!. 50 c:J Not shared sampled. During this period, 59 salmonid kills by resi :S2 dent killer whales in the area were documented, 95 % 0 40 (]) of which were chinook and 5 % coho (Fig. 4A). These 2 Ol co 30 species each represented less than 1 % of the available

(]) salmonids in the area. During 22 to 31 August 2004, the 20 (]) abundance of chinook increased to approximately 6 % of the available salmonids, and represented 97- % of 10 kills during this period (Fig. 4B). A substantial change 0 in the relative abundance of both salmonids and kills 2 3 4 5 6 Chinook age (yr) by killer whales took place between 12 and 18 October 2004 (Fig. 4C). Autumn-migrating chum salmon com Fig. 3. Oncorhynchus tshawytscha. Age distribution of chi prised 97- % of the test fishery catch during this period, nook salmon taken in shared (n = 117 kills) and non-shared (n = 22 kills) feeding events by Orcinus orca and chum was the predominant salmonid taken by feeding killer whales (90 % of kills). Chinook still rep resented 5 % of kills during this period despite being 100 extremely scarce in test catches (1 chinook in 57- 435 A 15 July-15 August 2004 salmonids sampled). 80 To evaluate whether resident killer whales preyed Test fishery equally on all sizes of their preferred prey species, the 60 □ □ Whale kills age distribution of chinook salmon taken by whales 40 was compared to the age distribution of chinook avail able to them for waters off NE Vancouver Island in 20 2000 to 2004. The relative abundance of chinook age classes was determined from estimates developed by 0 the Pacific Salmon Commission Joint Chinook Tech 100 nical Committee1. Although killer whales took all 5 yr B 22-31 August 2004 co classes available to them (Fig. 5), the frequency distri 80 .!!!. butions were significantly different, with killer whales taking fewer young chinook and more older chinook :S2 60 than expected based on the proportions of those ages (]) Ol 40 available. The mean age of chinook taken by whales co was 4.20 yr (±0.06 SE, n = 124), significantly older than (]) 20 the mean of 3.52 yr (±0.03 SE, n = 97-6,212) for avail able chinook (U-test = 33057-,p < 0.001). 0 100 C 12-18 October 2004 DISCUSSION 80 Assessing the selectivity of a predator requires infor 60 mation on both the predator's diet and the relative 40 availability of its various prey types. Although such information is often readily available for terrestrial 20 predators (e.g. Scheel 1993, Karanth & Sunquist 1995), it can be very difficult to obtain for fish-eating marine 0 Chinook Chum Coho Pink Sockeye mammals. The diet of aquatic predators must usually Species be determined by indirect means, such as from stom ach content, faecal, or fatty acid analyses, and the Fig. 4. Species composition of salmonids caught in seine test fisheries (open bars) and by resident killer whales (shaded bars) in western Johnstone Strait area, NE Vancouver Island from (A) 15 July to 15 August, (B) 22 to 31 August and (C) !Descriptions of the PSC chinook model and calibration proce 12 to 18 October 2004. Samples sizes were (A) 69 847 (test dures are provided in Pacific Salmon CommissionJoint Chi fishery) and 59 (whale kills). (B) 1548 (test fishery) and 30 nook Technical Committee Report TCChinook (97)-2, 1997, (whale kills), and (C) 57 435 (test fishery) and 41 (whale kills). and Report TCChinook (04)-2. Available at www.psc.org/ Specific names of prey in Table 2 pub!ications_tech_techcommitteereport.htm#TCCHINOOK

361 194 Mar Ecol Prog Ser 316: 185-199, 2006

362 Ford & Ellis: Selective foraging by killer whales 195

lieve that the scarcity of non-salmonids in our samples when migrating sockeye and pink salmon form the is a true reflection of their minor role in the diet of resi overwhelming majority of salmonids available to the dent killer whales, at least during the seasons covered whales (our Fig. 4; Groot & Margolis 1991, Nichol & by our observations. For example, lingcod Ophiodon Shackelton 1996, Wydoski & Whitney 2003, Quinn elongatus, found in stomach remains of a stranded res 2005). With only a single sockeye kill sample collected, ident whale (Ford et al. 1998), reach similar sizes to chi it is clear that despite its abundance during the sum nook salmon and are most abundant at depths of 10 to mer migratory period, this species is rare in the diet of 100 m, which overlap the preferred depths of chinook resident killer whales. Pink salmon were also very salmon (Cass et al. 1990, Candy and Quinn 1999). Ling uncommon in our samples, and the species does not cod have extremely small scales that are unlikely to be appear to be a significant prey item. Coho salmon are shed or recovered, but fish species identification is not relatively uncommon throughout the region (Groot reliant on scales alone. Tissue samples, from which spe and Margolis 1991, Wydoski and Whitney 2003, Rid cies identity can be readily determined using molecular dell 2004, Quinn 2005), and were consistently repre techniques, were collected from almost half of the feed sented in small numbers in killer whale feeding events. ing events sampled during 2003 to 2005. Of 21 prey Early summer runs of chum salmon occur during June items identified solely by DNA analysis of tissue frag and July, particularly on the northern coast of British ments, only 1, a sablefish, was a non-salmonid. We thus Columbia (Riddell 2004), and this species represented conclude that any negative bias in representation of a significant component of the killer whale diet in this bottomfishes in our samples is unlikely to be significant. period and region. During September and October, the Studies of diving behaviour in southern resident killer diet of northern resident killer whales shifts to predom whales found most activity to be concentrated in the inantly chum salmon, although chinook are also taken upper portion of the water column (Baird et al. 2005), frequently. This period coincides with the autumn which would be expected of a predator focused on migration of chum salmon through the area, which salmonid rather than demersal prey. starts in mid-September, peaks in mid-October, and is over by late October (Ryall et al. 1999). Comparisons of prey selection by resident killer Prey selection whale communities and clans revealed few differ ences. Northern and southern resident communities During the months of May to October, the main pe both fed predominantly on chinook. The greater pro riod of our field studies, salmonids are clearly the pre portion of chum in the samples of northern resident A ferred prey type of resident killer whales, representing and G clans probably resulted from their presence off over 96 % of identified prey. The only non-salmonids NE Vancouver Island during September and October found were a Pacific halibut, 2 rockfishes, a sablefish 2003 and 2004, when intensive sampling was under and 3 herring. Although the halibut and sablefish were taken. Little sampling of feeding events by R and J eaten, both rockfish species were abandoned by the clans during autumn has yet been undertaken. It ap whales after being partially consumed. We suspect that pears likely that southern resident whales also target rockfishes may be an undesirable prey type due to their chum salmon in addition to chinook in the autumn, as prominent dorsal spines, as this part of the fishes' body their movement into waters of Puget Sound in late was discarded. No rockfish remains have been found in October and November coincides with migratory stomach contents of stranded resident killer whales aggregations of these species (Osborne 1999). The sin (Ford et al. 1998). It seems unlikely that the herring gle sockeye and 2 steelhead salmon kills were ob were targeted prey items, since herring scales were col served only in southern residents, but this might be lected only during feeding sessions involving chinook due to chance, as a result of such small sample sizes. prey. As chinook feed extensively on herring (Healey Different age and sex classes of resident killer whales 1991), it is likely that herring scales were released did show some differences in salmonid prey composi when whales killed and broke chinook apart or were tion. Kills of pink salmon were made primarily by juve left in the water after chinook predation on herring. nile whales, and such predation of these small Chinook and chum salmon together represented salmonids may represent a form of play or practice 94 % of salmonids identified from kills by resident (Jacobsen 1986, Ford et al. 1998). Chinook, chum and whales. Chinook is the predominant species taken dur coho were taken by both adults and juveniles and by ing May to August. This would be expected for May both sexes, but a higher proportion of chum salmon and June, since other salmonids are uncommon in kills were recorded for adult males than for adult nearshore waters during these months (Groot & Mar females. As noted previously, this difference may be golis 1991, Quinn 2005). However, chinook remains due to a bias towards disproportionately greater sam the predominant prey species in July and August, pling of kills by males in September and October,

363 196 Mar Ecol Prog Ser 316: 185-199, 2006

when chum salmon was the predominant prey species. comm.). Killer whales in Prince William Sound and in Bain (1989) suggested that the larger body size of the Bering Sea are also known to remove and eat fishes males may allow them to dive more deeply and take from long line fishing gear, but they do so selectively, larger prey than females. Baird et al. (2005) observed according to energy value and size. Pacific halibut, male southern residents to dive deeper more fre sablefish and Greenland turbot Reinhardtius hippo quently than adult females. However, we observed no glossoides are among the favoured species, and the difference in the mean ages, and thus general sizes, of whales take the largest individuals of these species chinook taken by the 2 sexes. from fishing lines while ignoring other species such as Important factors influencing prey selection by pre Pacific cod Gadus macrocephalus, walleye pollock dators include size, energy density, availability, and Theragra chalcogramma, and rockfish Sebastes spp. catchability of prey (Stephens & Krebs 1986, Scheel (Matkin & Saulitis 1994, Yano & Dahlheim 1995). 1993, Lawson et al. 1998, Bowen et al. 2002). Resident Favoured species have higher average lipid content killer whales probably prefer chinook and chum and energy densities than species that are shunned salmon over other salmonids primarily because of their (Stansby 197-6, Winship & Trites 2003). large size. Chinook salmon are the largest of the sal Although size and lipid content may be important monids, and can attain masses of >25 kg (Healey 1991, factors in prey selection by resident killer whales, it is Quinn 2005). Most chinook taken by killer whales not clear why so few sockeye and pink salmon are were 4 to 6 yr old, representing mean masses of >8 kg. taken given their considerable seasonal abundance. Chum salmon are smaller, having mean masses of 5.0 During the peak of their migration through whale for to 7-.5kg (Ricker 1980, Salo 1991). Both species are sig aging areas, from mid-July to mid-September, sockeye nificantly larger than coho (mean mass= 2.95 kg; San and pink salmon combined outnumber chinook by as dercock 1991), sockeye (mean mass= 2.7-3kg; Burgner many as 500 to 1 (DFO seine test fisheries data, see 1991) and pink salmon (mean mass = 1.7- to 2.4 kg; 'Results'). We believe that it is a combination of the chi Heard 1991). Northern resident killer whales foraging nook salmon's large size, high lipid content, and year off NE Vancouver Island selected chinook that were round availability that makes them the salmonid of older on average than would be predicted from the age choice for resident whales in all seasons, even when distribution of available chinook, indicating a prefer alternative species are available in greater abundance ence for larger-sized fish of this species. It is interest during brief migratory pulses. ing, however, that the whales also took significant Unlike most salmonids, chinook are available to res numbers of 3 yr old chinook, which are similar in size ident killer whales in nearshore waters of the region to the far more abundant but rarely selected sockeye throughout the year. The 'ocean-type' chinook ranges and pink salmon. It may be that the relatively high over continental shelf waters throughout the marine lipid content of chinook compared to other salmonids portion of its life cycle (Healey 1991). The timing of (Stansby 197-6, Healey 1986, Winship & Trites 2003) is migration to spawning rivers is highly variable in chi another factor in the whales' preference for this spe nook, with different populations of chinook entering cies. However, chum salmon is a significant prey spe freshwater from mid-spring through autumn (Healey cies, despite having lower average lipid content than 1991). Migrating chinook also tend to travel through other salmonids (Stansby 197-6), suggesting that size coastal waters at slow rates of speed compared to other may be a more important influence than lipid content salmonids (Candy & Quinn 1999). Sockeye, chum and in determining salmonid preferences. pink salmon, on the other hand, have oceanic distribu Evidence from other regions also suggests that both tions most of their lives and only transit coastal regions prey size and lipid content may influence prey choice briefly while en route to spawning rivers (Groot & Mar by fish-eating killer whales. Saulitis et al. (2000) found golis 1991). The oceanic range of these salmonids is that resident killer whales in Prince William Sound, vast (Groot & Margolis 1991), with fishes distributed Alaska, appeared to prey selectively on coho salmon, widely at densities that are probably too low for whales which are larger and have higher lipid content than to effectively utilise. Although resident killer whales the far more abundant pink salmon that were available range extensively along the coast, there is no evidence to the whales during their study. Chinook are ex that they undertake long distance movements to off tremely rare in Prince William Sound during July and shore areas (Ford et al. 2000, Wiles 2004). Sockeye, August, when prey sampling was conducted by Sauli chum, and pink salmon, the 3 most abundant sal tis et al. (2000), which probably explains why few of monids in the North Pacific, are thus only available for this species were found in their samples. These whales predation by resident killer whales for a small portion prey extensively on chinook at other times and in other of the year compared to chinook salmon. areas where this prey species is present (C. Matkin, Killer whale populations tend to have specialised for North Gulf Oceanic Society, Homer, Alaska, pers. aging tactics to hunt their preferred prey (Lopez &

364 Ford & Ellis: Selective foraging by killer whales 197

Lopez 1985, Guinet 1992, Barrett-Lennard et al. 1996, A significant gap in our knowledge of the feeding Baird 2000, Saulitis et al. 2000, Pitman & Ensor 2003), ecology of resident killer whales is their diet during and this is likely to be the case for resident killer November to April. Chinook salmon remain available whales also. Resident killer whales may be particularly during these months in the inshore summer and skilled at finding and capturing chinook, which tend to autumn concentration areas of resident killer whales, travel more individually, at greater depths and closer but mostly at low densities (Osborne 1999; B. Riddell, to shore than smaller, schooling salmonids (Groot & Pacific Biological Station, Nanaimo, pers. comm.). Margolis 1991, Candy & Quinn 1999). The echoloca Whales disperse from these areas during winter and tion signals of resident killer whales are well suited for spring, but their whereabouts are for the most part the detection of echoes from individual chinook at unknown (Ford et al. 2000, Wiles 2004). Their prey ranges of 100 m or more (Au et al. 2004). Efficient pre may shift in winter and early spring to include more dation of smaller, schooling salmonids such as pink non-salmonid fishes, but there is little evidence to and sockeye salmon would probably require spe determine the extent to which their diet may change. cialised tactics such as the 'carousel' technique used The stomach of a northern resident whale that died in by killer whales to feed on herring in northern Norway late November off NE Vancouver Island contained (Simila & Ugarte 1993). Such coordinated foraging remains of chinook salmon and a variety of demersal behaviour has not been observed in resident killer fish species, including lingcod, sablefish, and green whales (Heimlich-Boran 1988, Ford 1989). It is note ling (Hexagrammossp.; Ford et al. 1998). It is probable worthy that resident killer whales do not target the that the resident killer whales' preference for chinook large seasonal concentrations of spawning herring that continues throughout the winter, and that they travel occur in coastal waters throughout their range. more widely over remote parts of the coast in pursuit of The known distribution patterns of resident killer this species. Future studies are needed to locate resi whales are consistent with what would be expected of dent whales during these months, and to determine a predator focused on chinook salmon. From May to whether chinook salmon is indeed their prey of choice early July, northern residents are found mostly along throughout the year. the coasts of the northern mainland of British Colum bia and the Queen Charlotte Islands, where they con Acknowledgements. The prey fragment sampling technique gregate in areas of high chinook density and feed pri central to this work was originally developed by the late marily on this species (Ford et al. 2000, authors' lvlichael A. Bigg, a pioneer of modern cetacean science. We unpubl. data). This period coincides with the earliest of thank the many research colleagues and volunteers who chinook runs in the region (Riddell 2004). Early-sum helped in the collection of data reported here, including R. Abernethy, K. Balcomb, L. Barrett-Lennard, J. Barrowman, mer-run chum salmon are also found there during this L. Dalla Rosa, M. and B. De Roos, J. Ellis, B. Falconer, B. Ford, time, although they are preyed upon to a lesser extent M. Ford, B. Gisborne, T. Hardy, J. Hildering, M. Janeway, than chinook. By mid-July, the whales begin to occur L. Spaven and J. Watson. For analysis of fish scales and tissue regularly off NE Vancouver Island, concurrently with remains, we thankS. MacLellan and staff of the Fish Ageing Laboratory, and J. Candy and K. Miller of the Molecular an increase in chinook abundance in the area and the Genetics Laboratory, Pacific Biological Station, Department of arrival of migrating sockeye and pink salmon. From Fisheries and Oceans, Nanaimo, British Columbia. Assistance mid-July to early September, when sockeye and pink with data preparation was provided by K. Davies and L. L. are transiting these waters, chinook salmon are also Spaven, and Nichol and P. Olesiuk helped with data analy sis and interpretation. We are also grateful to R. McNicol, B. available at relatively high densities. Although the Riddell and T. Quinn for providing information on chinook absolute abundance of chinook is far lower than that of abundance and for helpful discussions about salmonid life these smaller species, there are probably sufficient history and ecology. Earlier drafts of this manuscript bene fited from comments provided by D. Bowen, S. Ferguson, J. numbers available to meet the needs of resident L. whales without their having to switch to pink or sock Lawson, Nichol, G. Stenson, R. Stewart, D. Tollit, H. Yurk, J. Watson and 4 anonymous reviewers. This work was sup eye prey. By October, autumn-migrating chum salmon ported by Fisheries and Oceans Canada. The British Colum become the predominant salmonid available and the bia Wild Killer Whale Adoption Program, Vancouver Aquar principle prey of northern residents. Chinook salmon ium Marine Science Centre, Stubbs Island Whale Watching, appear to still be taken preferentially during this Telegraph Cove, and Langara Fishing Adventures, Vancou ver, provided valuable logistical support. period, since this species occurred more often in prey samples than would be predicted from its low relative abundance. The movements of southern residents to LITERATURE CITED foraging areas off southern Vancouver Island during Altmann J (1974) Observational study of behavior: sampling June to September also coincide with increased avail methods. Behaviour 49:227-265 ability of chinook salmon, and in October and Novem Au WWL, Ford JKB, Horne J, Newman Allman K (2004) ber with chum salmon (Osborne 1999). Echolocation signals of free-ranging killer whales (Orcinus

365 198 Mar Ecol Prog Ser 316: 185-199, 2006

orea) and modeling of foraging for chinook salmon (Onco Ford JKB (1989) Acoustic behaviour of resident killer whales rhynchus tshawytscha). J Acoust Soc Am 115:901-909 (Orcinus orca) off Vancouver Island, British Columbia. Bain DE (1989) An evaluation of evolutionary processes: stud Can J Zool 67:727-745 ies of natural selection, dispersal, and cultural evolution in Ford JKB, Ellis GM (1999) Transients: mammal-hunting killer killer whales (Orcinus orca). PhD dissertation, University whales of British Columbia, Washington, and Southeast of California, Santa Cruz ern Alaska. UBC Press, Vancouver Baird RW (2000) The killer whale: foraging specializations Ford JKB, Ellis GM, Barrett-Lennard LG, Morton AB, Palm and group hunting. In: Mann J, Connor RC, Tyack PL, RS, Balcomb KC Ill (1998) Dietary specialization in two Whitehead H (eds) Cetacean societies: field studies of dol sympatric populations of killer whales (Orcinus orca) in phins and whales. University of Chicago Press, Chicago, coastal British Columbia and adjacent waters. Can J Zool IL, p 127-153 76:1456-1471 Baird RW, Dill LM (1995) Occurrence and behaviour of tran Ford JKB, Ellis GM, Balcomb KC (2000) Killer whales: the nat sient killer whales: seasonal and pod-specific variability, ural history and genealogy of Orcinus orca in the waters of foraging, and prey handling. Can J Zool 73:1300-1311 British Columbia and Washington, 2nd edn. UBC Press, Baird RW, Hanson MB, Dill LM (2005) Factors influencing the Vancouver diving behaviour of fish-eating killer whales: sex differ Ford JKB, Ellis GM, Matkin DR, Balcomb KC, Briggs D, ences and die! and interannual variation in diving rates. Morton AB (2005) Killer whale attacks on minke whales: Can J Zool 83:257-267 prey capture and antipredator tactics. Mar Mamm Sci 21: Barrett-Lennard LG (2000) Population structure and mating 60 3-618 patterns of killer whales, Orcinus orca, as revealed by Groot C, Margolis L (eds) (1991) Pacific salmonlife histories. DNA analysis. PhD dissertation, University of British UBC Press, Vancouver Columbia, Vancouver Guinet C (1990) Sympatriedes deux categories d'orques dans Barrett-Lennard LG, Ford JKB, Heise KA (1996) Themixed bless le detro it de Johnstone, Columbie Britannique. Rev Ecol ing of echolocation: differences in sonar use by fish-eating Terre Vie 45:25-34 and mammal-eating killer whales. AnirnBehav 51:55 3-565 Guinet C (1992) Comportement de chasse des orques (Orci Bigg MA (1982) An assessment of killer whale (Orcinus orca) nus orca) autour des iles Crozet. Can J Zool 70:1656-1667 stocks off Vancouver Island, British Columbia. Rep Int Guinet C, Bouvier J ( 1995) Development of intentional strand Whal Comm 32:655-666 ing hunting techniques in killer whale (Orcinus orca) Bigg MA, Ellis GM, Ford JKB, Balcomb KC (1987) Killer calves at Crozet Archipelago. Can J Zoo! 73:27-33 whales: a study of their identification, genealogy and nat Haley D (ed) (1986) Marine mammals of eastern North Pacific ural history in British Columbia and Washington State. and Arctic waters, 2nd edn. Pacific Search Press, Seattle, Phantom Press, Nanairno WA Bigg MA, Olesiuk PF, Ellis GM, Ford JKB, Balcomb KC Ill Healey MC (1986) Optimum size and age at maturity in (1990) Social organizationand genealogy of resident killer Pacific salmon and effects of size-selective fisheries. In: whales (Orcinus orca) in the coastal waters of British Meerburg DJ (ed) Salmonid age at maturity. Department Columbia and Washington State. Rep Int Whal Comm of Fisheries and Oceans, Ottawa, p 39-52 Spec Issue12:383-405 Healey MC (1991) Life history of chinook salmon (Onco Bowen WD (1997) Role of marine mammalsin aquatic ecosys rhynchus tshawytscha). In: Groot C, Margolis L (eds) tems. Mar Ecol Prog Ser 15 8:267-274 Pacific salmon life histories. UBC Press, Vancouver, Bowen WD, Siniff DB (1999) Distribution, population biology, p 313-393 and feeding ecology of marine mammals. In: Reynolds JEI, Heard WR (1991) Life history of pink salmon (Oncorhynchus Rommel SA (eds) Biology of marine mammals. Smithson gorbuscha). In: Groot C, Margolis L (eds) Pacific salmon ian Press, Washington, DC, p 423-4 84 life histories. UBC Press, Vancouver, p 12 1-230 Bowen WD, Tully D, Boness DJ, Bulheier BM, Marshall GJ Heimlich-Boran J (1986) Fishery correlations with the occur (2002) Prey-dependent foraging tactics and prey profitabil rence of killer whales in greater Puget Sound. In: ity in a marine mammal. Mar Ecol Prog Ser 244:235-245 Kirkevold BC, Lockard JS (eds) Behavioral biology of Burgner RL (1991) Life history of sockeye salmon (Onco killer whales. Alan R. Liss, New York, p 113-131 rhynchus nerka). In: Groot C, Margolis L (eds) Pacific Heimlich-Boran JR (1988) Behavioral ecology of killer whales salmon life histories. UBC Press, Vancouver, p 3-117 ( Orcinus orca) in the Pacific Northwest. Can J Zoo! 66: Candy JR, Quinn TP (1999) Behaviour of adult chinook 565-578 salmon (Oncorhynchus tshawytscha) in British Columbia Hoelzel AR (1993) Foraging behaviour and social group coastalwaters determined from ultrasonic telemetry. Can dynamics in Puget Sound killer whales. Anirn Behav 45: J Zool 77:1161-116 9 581-591 Cass AJ, Beamish RJ, Mcfarlane GA (1990) Lingcod Hoelzel AR, Dahlheim M, Stern SJ (1998) Low genetic varia (Ophiodon elongatus). Can Spec Pub! Fish Aquat Sci 10 9 tion among killer whales (Orcinus orca) in the eastern DeMaster DP, Trites AW, Clapham PJ, Mizroch SA, Wade PR, North Pacific, and genetic differentiation between forag Small RJ, Ver Hoef JM (2006) The sequential megafaunal ing specialists. J Hered 89:12 1-128 collapse hypothesis: testing with existing data. Prog Jacobsen JK (1986) The behavior of Orcinus orca in the John Oceanogr 68:329-342 stone Strait, British Columbia. In: Kirkevold BC, Lockard Felleman FL, Heimlich-Boran JR, Osborne RW (1991) The JS (eds) Behavioral biology of killer whales. Alan R. Liss, feeding ecology of killer whales (Orcinus orca) in the New York, p 135-185 Pacific northwest. In: Pryor K, Norris KS (eds) Dolphin Jefferson TA, Stacey PF, Baird RW (1991) A review of killer Societies: discoveries and puzzles. University of California whale interactions with other marine mammals: predation Press, Berkeley, p 113-147 to co-existence. Mammal Rev 21:151-180 Fertl D, Acevedo-Gutierrez A, Darby FL (1996) A report of Karanth KU, Sunquist ME (1995) Prey selection by tiger, killer whales (Orcinus orca) feeding on a carcharhinid leopard and dhole in tropical forests. J Anim Ecol 64: shark in Costa Rica. Mar Mamm Sci 12:606-611 439-450

366 Ford & Ellis: Selective foraging by killer whales 199

Kuhn BR (1988) TheMRP-reporter program: a data extraction work management strategy. CSAS Rep No. 99/169, and reporting tool for the mark recovery program data Fisheries and Oceans Canada, Ottawa (also available at base. Can Tech Rep Fish Aquat Sci 1625 www.dfo-mpo.gc.ca/CSAS/) Lawson JW, Anderson JT, Dalley EL, Stenson GB (1998) Salo EO (1991) Life history of chum salmon (Oncorhynchus Selective foraging by harp seals Phoca groenlandica in keta). In: Groot C, MargolisL (eds) Pacific salmonlife his nearshore and offshore waters of Newfoundland, 1993 tories. UBC Press, Vancouver, p 233-309 and 1994. Mar Ecol Prog Ser 163:1-10 Sandercock FK (1991) Life history of coho salmon (Oncorhyn Lopez JC, Lopez D (1985) Killer whales (Orcinus orca) of chus kisutch). In: Groot C, Margolis L (eds) Pacific salmon Patagonia, and their behavior of intentional stranding life histories. UBC Press, Vancouver, p 397-445 while hunting nearshore.J Mammal 66:181-183 Saulitis E, Matkin C, Barrett-Lennard L, Heise K, Ellis G Mann J (1999) Behavioral sampling methods for cetaceans: a (2000) Foraging strategies of sympatric killer whale (Orci review and critique. MarMamm Sci 15:102-122 nus orca) populations in Prince William Sound, Alaska. Martinez DR, Klinghammer E (1970) The behavior of the Mar Mamm Sci 16:94-109 whales, Orcinus orca: a review of the literature.Z Tierpsy Scheel D (1993) Profitability, encounter rates, and prey choice chol 27:828-839 of African lions. Behav Ecol 4:90-97 Matkin CO, Saulitis EL (1994) Killer whale ( Orcinus orca) Simila T, Ugarte F (1993) Surface and underwater observa biology and management in Alaska. Contract No. tions of cooperatively feeding killer whales in northern T75135023, Marine Mammal Commission, Washington, Norway. Can J Zool 71:1494-1499 DC Simila T, Holst JC, Christensen I (1996) Occurrence and diet Mizroch SA, Rice DW (2006) Have North Pacifickiller whales of killer whales in northern Norway: seasonal patterns rel switchedprey species in response to depletion of the great ative to the distribution and abundance of Norwegian whale populations? MarEcol Prog Ser 310:235-246 spring-spawning herring. Can J Fish Aquat Sci 53: Nichol LM, Shackleton DM (1996) Seasonal movements and 769-'i"i'9 foraging behaviour of northern resident killer whales SpringerAM, Estes JA, van Vliet GB, Williams TM, Doak DF, (Orcinus orca) in relation to the inshore distribution of Danner EM, Forney KA, Phister B (2003) Sequential salmon (Oncorhynchus spp.) in British Columbia. Can megafaunal collapse in the North Pacific Ocean: an ongo J Zoo! 74:983-991 ing legacy of industrial whaling? Proc Natl Acad Sci USA Olesiuk PF, Bigg MA, Ellis GM (1990) Life history and popu 100: 12223-12228 lation dynamics of resident killer whales ( Orcinus orea) in Stansby ME (1976) Chemical characteristics of fish caught in the coastal waters of British Columbia and Washington the northeast Pacific Ocean. Mar Fish Rev 38:1-11 State. Rep Int Whal Comm Spec Issue 12:209-242 Stephens DW, Krebs JR (1986) Foraging theory. Princeton Osborne RW (1999) A historical ecology of Salish Sea 'resi University Press, Princeton, NJ dent' killer whales (Orcinus orca): with implications for Tollit, D, Greenstreet SPR, ThompsonPM (1997) Prey selec management. PhD dissertation, University of Victoria, Vic tion by harbour seals, Phoca vitulina, in relation to varia toria tions in prey abundance. Can J Zoo! 75:1508-1518 Pitman RL, Ensor P (2003) Three forms of killer whales (Orci Wiles GJ (2004) Washington State status report for the killer nus orca) in Antarctic waters. J Cetacean Res Manag 5: whales. Washington Department of Fish and Wildlife, 131-139 Olympia, WA (also available at http://wdfw.wa.gov/) Quinn TP (2005) The behavior and ecology of Pacific salmon Winship AJ, Trites AW (2003) Prey consumption of Steller sea and trout. UBC Press, Vancouver lions (Eumetopias jubatus) off Alaska: how much prey do Rendell LE, Whitehead H (2001) Culture in whales and dol they require? Fish Bull (Wash DC) 101 :147-163 phins. Behav Brain Sci 24:309-331 Withler RE, Candy JR, Beacham TD, Miller KM (2004) Fo Ricker WE (1980) Changes in the age and size of chum rensic DNA analysis of Pacific salmonid samples for salmon (Oncorhynchus keta). Can Tech Rep Fish Aquat species and stock identification. Environ Biol Fish 69: Sci 930 275-285 Riddell B (2004) Pacific salmon resources in central and Wydoski RS, Whitney RR (2003) Inland fishes of Washington. north coast British Columbia, Pacific Fisheries Resource University of Washington Press, Seattle Conservation Council, Vancouver (also available at Yano K, Dahlheim ME (1995) Behavior of killer whales Orci www.fish.bc.ca/) nus orca duringlongline fishery interactions in the south Ryal! P, MurrayC, Palermo V, Bailey D, Chen D (1999) Status eastern Bering Sea and adjacent waters. Fish Sci 61: of clockwork chum salmon stock and review of the clock- 584-589

Editorialresponsibility: Howard I. Browman (Associate Submitted: September 12, 2005; Accepted: November 15, 2005 Editor-in-Chief), Storebe, Norway Proofs received from author(s): June 16, 2006

367

C S A S S C C S

Canadian Science Advisory Secretariat Secrétariat canadien de consultation scientifique

Research Document 2009/101 Document de recherche 2009/101

Chinook salmon predation by Prédation du saumon quinnat par les resident killer whales: épaulards résidents : sélectivité seasonal and regional selectivity, saisonnière et régionale, identité des stock identity of prey, and stocks de proies et taux de consumption rates consommation

John K.B. Ford Brianna M. Wright Graeme M. Ellis John R. Candy

Fisheries and Oceans Canada, Pacific Biological Station 3190 Hammond Bay Road, Nanaimo, BC V9T 6N7 CANADA

This series documents the scientific basis for the La présente série documente les fondements evaluation of aquatic resources and ecosystems in scientifiques des évaluations des ressources et Canada. As such, it addresses the issues of the des écosystèmes aquatiques du Canada. Elle day in the time frames required and the traite des problèmes courants selon les documents it contains are not intended as échéanciers dictés. Les documents qu’elle definitive statements on the subjects addressed contient ne doivent pas être considérés comme but rather as progress reports on ongoing des énoncés définitifs sur les sujets traités, mais investigations. plutôt comme des rapports d’étape sur les études en cours.

Research documents are produced in the official Les documents de recherche sont publiés dans language in which they are provided to the la langue officielle utilisée dans le manuscrit Secretariat. envoyé au Secrétariat.

This document is available on the Internet at: Ce document est disponible sur l’Internet à: http://www.dfo-mpo.gc.ca/csas/

ISSN 1499-3848 (Printed / Imprimé) ISSN 1919-5044 (Online / En ligne) © Her Majesty the Queen in Right of Canada, 2010 © Sa Majesté la Reine du Chef du Canada, 2010

368

369

Correct citation for this publication:

Ford, J.K.B, Wright, B.M., Ellis, G.M., and Candy, J.R. 2010. Chinook salmon predation by resident killer whales: seasonal and regional selectivity, stock identity of prey, and consumption rates. DFO Can. Sci. Advis. Sec. Res. Doc. 2009/101. iv + 43 p.

ABSTRACT

Resident killer whales (Orcinus orca) found in coastal waters of the cold-temperate northeastern Pacific are fish-feeding predators that specialize on Pacific salmon. Field studies have shown that although most available salmonids are consumed, Chinook salmon (Oncorhynchus tshawytscha) is the whales’ primary prey species, most likely because of its large size, high lipid content, and year-round occurrence in coastal waters. Chinook salmon availability appears to be important to the survival and recovery of resident killer whale populations. In this report we describe the results of recent field studies and analyses aimed at improving our understanding of the role played by Chinook salmon in the seasonal foraging ecology and energetics of resident killer whales. An additional 410 prey items identified from scale and tissue samples collected at the sites of resident feeding events provide further support for the importance of Chinook salmon in most seasons and coastal areas. Genetic stock identification of prey samples indicate that killer whales feed on Chinook salmon originating from a variety of regions between Southeast Alaska and Oregon, with stocks in the Fraser River system being of particular importance both coast-wide and in Critical Habitats. An updated analysis confirms the long-term correlation between survival of resident killer whales and range-wide Chinook abundance, though recent declines in Chinook abundance have not yet been associated with increased mortality rates. Estimates of Chinook salmon consumption based on daily prey energy requirements and diet composition suggest that resident killer whale populations at their current abundance may require over 1,000,000 Chinook per year, roughly equivalent to recent annual levels of harvests of this species in commercial and recreational marine fisheries. Estimates of Chinook salmon requirements for northern and southern resident killer whale populations in their Critical Habitats are also provided, as is an estimate of the Chinook abundance that would be required to support killer whale recovery over the next decade. Although the information in this report may be useful for future conservation and management of resident killer whales and their primary prey, further studies are needed to resolve existing uncertainties about year-round diet composition and feeding rates.

iii 370

RÉSUMÉ

Les épaulards résidents (Orcinus orca) des eaux côtières de la zone froide tempérée du nord- est du Pacifique sont des prédateurs piscivores qui préfèrent le saumon du Pacifique. Des études sur le terrain ont démontré que même si la plupart des salmonidés disponibles sont consommés, le saumon quinnat (Oncorhynchus tshawytscha) est la principale espèce proie des épaulards, fort probablement en raison de sa grande taille, de sa teneur élevée en lipides et de sa présence dans les eaux côtières toute l’année. La disponibilité de saumon quinnat semble importante pour la survie et le rétablissement des populations d’épaulards résidents. Dans ce rapport, nous décrivons les résultats des études sur le terrain et des analyses récentes visant à améliorer notre compréhension du rôle que joue le saumon quinnat pour l’écologie alimentaire saisonnière et l’énergétique des épaulards résidents. Les 410 proies supplémentaires identifiées à partir des restes d’écailles et de tissus recueillis sur les sites d’activités d’alimentation des épaulards résidents soulignent encore plus l’importance du saumon quinnat pour la plupart des régions côtières et des saisons. L’identification du stock génétique des restes de proies indique que les épaulards résidents se nourrissent de saumon quinnat provenant de diverses régions situées entre le sud-est de l’Alaska et l’Oregon, les stocks du système du fleuve Fraser étant particulièrement importants à la fois sur toute la côte et dans les habitats critiques. Une analyse actualisée confirme la corrélation à long terme entre le taux de survie des épaulards résidents et l’abondance de saumon quinnat pour toute l’aire, malgré le fait que la récente diminution d’abondance du saumon quinnat n’a pas encore été associée avec les taux de mortalité accrus. Les estimations de consommation du saumon quinnat en fonction des besoins énergétiques quotidiens en proies et de la composition de la diète portent à croire que les populations d’épaulards résidents, selon leur abondance actuelle, peuvent nécessiter un million de saumons quinnat par année, soit approximativement l’équivalent des récents niveaux annuels de capture des pêches marines commerciales et récréatives pour cette espèce. On donne aussi des estimations des besoins en saumon quinnat pour les populations d’épaulards résidents du nord et du sud dans leurs habitats critiques, de même qu’une estimation de l’abondance de saumon quinnat qui serait nécessaire pour favoriser le rétablissement de l’épaulard au cours de la prochaine décennie. Même si l’information donnée dans ce rapport peut être utile pour la conservation et la gestion futures des épaulards résidents et de leur principale proie, d’autres études sont nécessaires pour trouver une réponse aux incertitudes actuelles sur la composition de la diète et le taux de consommation toute l’année.

iv 371

INTRODUCTION

So-called ‘resident’ killer whales are one of three sympatric ecotypes of Orcinus orca found in the coastal waters of the cold-temperate northeastern Pacific. Each ecotype is ecologically specialized and has a distinct diet. Residents are fish feeders, and in particular specialize on Pacific salmon (Ford et al. 1998). The so-called ‘transient’ killer whale ecotype is a mammal-hunting specialist, feeding on pinnipeds and small cetaceans but not fish (Ford et al. 1998). The so-called ‘offshore’ killer whale ecotype is a poorly-known fish feeder found primarily on the outer coast and may specialize on sharks and other large fish (Ford et al. 2000; Dahlheim et al. 2008; Cetacean Research Program, Pacific Biological Station, unpubl. data).

Resident killer whales in British Columbia and adjacent coastal waters of Washington State have been the focus of annual field studies since the early 1970s (Bigg et al. 1976; Bigg 1982). A key method in these long-term studies has been photographic identification of individual whales using natural markings on the dorsal fin and back. This work has provided a great deal of information on the social organization, life history, population dynamics, and behavioural acoustics of these whales (e.g., Bigg 1982; Bigg et al. 1990; Olesiuk et al. 1990, 2005; Ford 1991; Ford et al. 2000).

Knowledge of the foraging ecology of resident killer whales has taken somewhat longer to acquire, mostly due to the difficulties in studying feeding behaviour in wild cetaceans. The association between seasonal aggregations of resident killer whales in inshore waters near Vancouver Island and the spawning migration of Pacific salmon has long suggested that that these whales feed extensively on this prey type (Heimlich-Boran 1986; Guinet 1990; Nichol and Shackleton 1996), but the first detailed evidence of salmonid consumption was documented by Ford et al. (1998). This study investigated diet by observing predation events, examining stomach contents of stranded whales and, in particular, opportunistically collecting fish scales and other prey fragments from kill sites during 1973-96. An unexpected finding of this research was that despite feeding on most available salmonid species, resident killer whales appeared to exhibit a strong preference for Chinook salmon (Oncorhynchus tshawytscha), the largest and most energy-rich salmonid in the region, but also one of the least common. The far more seasonally abundant pink (O. gorbuscha) and sockeye salmon (O. nerka) appeared not to be significant in the whales’ diet.

Although stomach contents analysis supported field observations that resident killer whales may feed preferentially on Chinook salmon, we had concerns that the prey fragment sampling technique used to identify prey species was biased in favour of large fish such as Chinook salmon (Ford et al. 1998). It seemed possible that Chinook, being larger than other salmonids, were more prone to being broken up prior to being eaten, thus shedding more scales in the process (Ford et al. 1998). However, without knowledge of the details of prey handling and consumption of salmonids and other fish species by resident killer whales, it was not possible to evaluate the significance, if any, of this potential bias. It was concluded that resident whales have a preference for Chinook, but the extent of this preference remained uncertain (Ford et al. 1998).

To address these concerns and to generally improve understanding of the diet of resident killer whales, we undertook dedicated field studies of foraging behaviour, including focal animal observations to document the details of prey capture, handling and consumption, during 2002-04. This research revealed that most salmonid feeding events involved sharing of prey among animals within the group, and that scales and tissue fragments were shed when fish were broken apart for this purpose (Ford and Ellis 2005, 2006). All salmonid species and

1 372

sizes appeared to be shared, which suggested that any bias in prey sampling was likely to be minimal, at least with respect to salmonid prey. This study provided strong support that Chinook salmon is the primary prey species of residents, and that the smaller pink and sockeye salmon were not significant components of the whales’ diet. It also revealed that chum salmon (O. keta) was an important prey species during late September and October.

Chinook salmon may be of such importance as primary prey of resident killer whales that its availability plays a role in their population dynamics. Ford et al. (2005, 2010) showed that mortality rates of resident killer whales were negatively correlated with Chinook salmon abundance over a 25-year period, from 1979-2003. In particular, a sharp decline in Chinook abundance during the late 1990s was associated with killer whale mortality rates up to 2-3 times greater than expected, which resulted in population declines in both resident killer whale populations, the so-called northern and southern residents. Calving rates showed a weaker, but still significant, positive correlation with Chinook salmon abundance. Ward et al. (2009) also found a significant association between Chinook abundance and reproductive rates in the southern resident population.

Resident killer whales are listed under the Species at Risk Act in Canada, with the northern population designated as threatened and the southern population as endangered (Fisheries and Oceans Canada 2008). The southern resident population is similarly listed as endangered under the U.S. Endangered Species Act (NMFS 2005). A primary objective in the Recovery Strategy for resident killer whales is to Ensure that resident killer whales have an adequate and accessible food supply to allow recovery (Fisheries and Oceans Canada 2008). Since Chinook salmon is the primary prey of resident killer whales and its abundance may directly affect survival and recovery, it is important that an improved understanding of the seasonal and geographic importance of this prey resource is obtained. In particular, information is needed on the specific Chinook salmon populations that are exploited by resident killer whales at different times of the year and in different parts of their range, and the overall abundance of Chinook that may be needed to support the existing resident killer whale population and to provide for sustained growth and recovery.

In this report, we describe new information and analyses on a variety of aspects of the dynamics between resident killer whales and their primary prey, Chinook salmon. First, we build on our understanding of diet composition described in Ford and Ellis (2005, 2006) by presenting the results of new prey sampling undertaken in 2005-09. We then describe the population identity of Chinook salmon sampled from resident killer whale feeding events to assess the regional importance of different stocks to foraging whales. Next, we investigate the effects of variations in Chinook salmon availability by updating the Ford et al. (2005) analysis of the relationship between resident killer whale mortality and Chinook abundance, and by examining potential prey shifts during periods of high and low Chinook abundance. Finally, we estimate the numbers of Chinook salmon that may be consumed by resident killer whales based on diet composition and energetic requirements, both annually and within designated Critical Habitat, and predict the abundance of Chinook salmon that may be needed to allow for population recovery over the next decade.

2 373

STUDY POPULATIONS AND DATA COLLECTION

Studies were undertaken in coastal waters of British Columbia during 1973-2009, primarily in nearshore waters off eastern and southwestern Vancouver Island, the central and northern mainland coast, and near Langara Island off the northwest coast of Graham Island. Two populations of resident killer whales, northern residents and southern residents, can be found in these waters in all months of the year, but mostly during May-November. The northern resident population is found mostly from mid Vancouver Island to southeastern Alaska, and the southern resident population off the southern half of Vancouver Island and in the inland waters of Washington state (Figure 1). Whales from the two populations have not been seen to associate despite extensive overlap in their ranges (Ford et al. 2000). Large aggregations of residents can be found in certain coastal locations during summer. Residents greatly reduce the use of these locations in winter and spring, and their range during this period is poorly known but is suspected to be more extensive in outer coast waters (Ford et al. 2000; Wiles 2004).

The northern and southern resident populations contained 86 and 252 individuals, respectively, in 2008 (Figure 2). Both populations have grown since first censused in 1974. The northern population has more than doubled in size, from 122 to 252 animals, while the southern population is only 21% larger (71 in 1974, 86 in 2008). Both populations grew at an overall annual rate of 2-3% from the early 1970s to the mid 1990s, and then experienced a decline in abundance in the late 1990s. This decline was driven by a sharp increase in mortality rates and, to a lesser extent, decreased recruitment (Ford et al. 2005; Olesiuk et al. 2005). The southern resident population has yet to recover from this period of decline, but the northern resident population is currently at its greatest abundance since the study began.

FIELD EFFORT AND PROCEDURES

Data on predation by resident killer whales have been collected annually since 1973, as part of long-term studies on the life history, social organization, acoustic behaviour, and population genetics of these animals (Bigg 1982; Bigg et al. 1990; Olesiuk et al. 1990, 2005; Ford 1989, 1991; Ford et al. 1998, 2000; Barrett-Lennard 2000). Data collected during 1973- 2002 consisted mostly of surface observations of feeding events and opportunistic collection of prey fragments from the vicinity of kills. Effort varied widely according to changing research objectives, but predation studies were given higher priority during 1990-2002 (Ford et al. 1998, Ford and Ellis 2005). In 2003-09, field studies were dedicated to systematically documenting foraging behaviour and collecting predation data, in addition to conducting the annual census of individuals by photo-identification (Bigg et al. 1987; Ford et al. 2000).

Field studies in 1974-2002 were conducted using a variety of vessels from 5-20 m in length. In 2003-09, dedicated studies of resident killer whale feeding were undertaken mainly from a 10-m long command-bridge power vessel. When whales were encountered, individuals were observed visually or photographed to determine identity from natural markings on the dorsal fin and back. Photographic identification procedures are described in Bigg et al. (1987) and Ford et al. (2000). Once the identity of killer whales present in the encounter was established, effort was directed to documenting foraging behaviour and collecting scales and tissue fragments from prey killed during feeding events. Activity state of the whales was determined from surfacing and dispersion patterns (see Ford 1989 for definitions of activity states). When foraging, whale groups typically spread out over several square kilometres, with individuals and subgroups swimming and diving independently but travelling generally in the same direction. Surfacing individuals and groups were scanned by eye or binoculars for signs

3 374

of prey pursuit or capture. When apparent feeding was observed, the site of the kill was approached promptly (while taking care to avoid disturbing the animals) in order to determine or confirm identities of the whale(s) involved and to search for prey fragments in the water. Whether or not prey fragments were found, the individual or subgroup was then followed at distances of 25-50 m to document subsequent feeding events. These focal individual and subgroup follows (Altmann 1974; Mann 1999) were maintained for as long as the whale(s) continued active foraging. Focal follows were terminated when animals joined other groups and could no longer be followed individually, when subgroups merged or split, or when other circumstances necessitated ending the session.

Focal individuals and subgroups were monitored closely and constantly during feeding sessions. Particular attention was given to direction of travel, regularity of dive durations, and degree of subgroup cohesion, as changes in these variables often signalled a feeding event. Individuals or subgroups suspected to have captured a prey item were approached to within 10 m to observe prey handling and consumption. To collect evidence of feeding, the surfacing locations of the feeding whale or subgroup were also examined for prey fragments at the surface or in the water column. The boat driver was positioned approximately 4 m above the water surface on the command bridge of the study vessel, which afforded a high-angle view into the water as the boat was manoeuvred. A second observer stood on the vessel’s bow, holding a fine-mesh dip net (mesh size approximately 1 mm) with 5-m telescoping handle, and also searched for fragments. When fish scales or bits of tissue were seen, the boat was immediately stopped and the net was deployed to retrieve the fragments. Fragments were collected mostly at depths of 0-2 m, but occasionally as deep as 3-4 m in calm conditions with good water clarity. Rain, winds greater than 10 kts, and high water turbidity reduced the success rate of fragment location and collection.

When prey fragments were collected, they were placed immediately in a 5 ml vial containing 95% ethanol. The date, time, and geographical position (from a differential GPS instrument) of the feeding event was recorded, as well as the identity of the individual making the kill and others involved in the prey capture or consumption.

PREY SPECIES IDENTIFICATION AND AGEING

Many species of fishes are readily identifiable at a distance by an experienced observer, but salmon species can be difficult to distinguish without close examination. Although Ford et al. (1998) included salmonid identifications based on field observations, in Ford and Ellis (2005) and the current analyses we chose to include only positive salmonid identifications based on scales or tissue samples to eliminate this potential source of error. Fish scales were analyzed by the Sclerochronology Laboratory at the Pacific Biological Station (Department of Fisheries and Oceans, Nanaimo, B.C.) to determine species identity and age according to procedures outlined in MacLellan (2004). Age was designated using the European method and age class was assigned according to the internationally-accepted January 1st birthdate. Species identification was based on diagnostic scale characteristics (MacLellan 2004).

Scales that could not be positively identified to species and tissue samples collected from feeding events were submitted to the Molecular Genetics Laboratory at the Pacific Biological Station for species identification using allelic size range of genomic DNA. Variation at twelve microsatellite loci were used to identify species as well as assign individual Chinook salmon to region of origin using a mixture analysis program cbayes (Neaves et al. 2005). The baseline consisted of 268 populations ranging from south-east Alaska to California using methodology reported in Beacham et al. (2003, 2006).

4 375

RESULTS

DIET OF RESIDENT KILLER WHALES BY SEASON AND REGION

Ford and Ellis (2005) described results of observations of predation and prey species identification from field studies of resident killer whales conducted during 1974-2004. They provided results of analyses of 487 feeding events documented during 197 encounters with resident killer whales over the 30 year study period. Sixty percent of these feeding events were documented in the last two years of this time series, reflecting a shift in research focus to foraging behaviour and diet. Feeding events were recorded during May to December, but none in January to April.

Continued field studies of resident killer whale diet since 2004 have almost doubled the dataset used in the Ford and Ellis (2005) analysis, and have broadened sampling both geographically and seasonally. The total dataset now available, and which forms the basis of the current analysis, includes 937 feeding events documented during 341 encounters with northern (n = 715 events) and southern (n = 222 events) resident killer whales. Feeding events were observed in all years between 1973 and 2009, although 81% of events were documented during dedicated studies of foraging since 2000 (Figure 3). Evidence used to identify prey for the 937 feeding events is provided in Table 1. Over 90% of feeding events were documented by collection and analysis of fish scales, tissue, or both.

The monthly distribution of feeding events for northern and southern resident killer whales is shown in Figure 4. Feeding was documented in all months except April, with 80% of events sampled during the summer months of June through September. The overall locations of feeding events involving northern and southern residents are depicted in Figure 5. To facilitate regional comparisons of diet composition, the coastal waters of British Columbia were divided into 6 regions, shown with the Pacific Fisheries Management Areas (PFMA) they encompass in Figure 6. Sample sizes of feeding events and prey species composition for each of these regions are tabulated in Table 2.

Prey species identification from this more extensive dataset provides further support for the conclusions reached by Ford and Ellis (2005, 2006). All feeding events involved fish, at least 97.4% of which were salmonids (Table 2). Twenty-four samples (2.5%) could not be identified to species, and some of these may also have included salmonids. The only non- salmonids identified were 2 Pacific herring (Clupea pallasi), 4 Pacific sardine (Sardinops sagax), 1 yelloweye rockfish (Sebastes ruberrimus), 1 quillback rockfish (Sebastes maliger), 1 sablefish (Anoplopoma fimbria), and 1 Pacific halibut (Hippoglossus stenolepis).

Chinook salmon was by far the predominant salmonid species consumed, representing 71% of the 806 salmonid kills identified to species. Chum salmon was second in importance at 24%, and coho, pink, sockeye and steelhead salmon were minor components of the whales’ diet at less than 3% each. The monthly distribution of salmonid species consumed is tabulated in Table 3 and shown graphically in Figure 7. Chinook salmon was the predominant prey species in all months except October and November, when chum salmon were consumed more frequently. Sample sizes for winter months are small, but Chinook salmon was the only prey species documented in January-March. Chinook salmon was also the predominant prey species in all coastal regions (Figure 8, Table 2).

5 376

Table 4 presents the ages of salmon consumed by resident killer whales as determined from scale samples recovered from feeding events. Of the 431 Chinook salmon samples aged, 325 (75%) were ‘ocean’ type fish as indicated by 0 years in fresh water, and 106 (25%) were ‘stream’ type fish that spent 1-2 years in fresh water prior to entering the sea. Stream type Chinook tend to migrate to the open ocean and only return to coastal waters during their spawning migration (Healey 1991). Ocean type Chinook, on the other hand, spend their entire life cycle within continental shelf waters and may thus be more available to the whales throughout the year.

POPULATION IDENTITY OF CHINOOK SALMON CONSUMED BY RESIDENT KILLER WHALES

In order to assess the seasonal and regional importance of different Chinook salmon populations in the diet of resident killer whales, we undertook genetic stock identification analyses of scale and tissue samples collected from feeding events. A total of 474 prey samples of Chinook salmon resulted in DNA suitable for stock region identification (Beacham et al. 2006). Chinook salmon prey of killer whales originated in 19 of the 38 regional stocks described in Beacham et al. (2006). Chinook from stock regions within the Fraser River system comprised 58% of samples. Also important were stocks in the east coast (10%) and west coast (8%) of Vancouver Island regions. In the following subsections, we present stock identities for Chinook sampled in the different coastal regions indicated in Figure 6.

Queen Charlotte Islands – PFMA 1-2

Locations and months of Chinook salmon sampled from feeding events and stock region identifications are shown in Figure 9. Thirty-three samples were collected from the north coast of the Queen Charlotte Islands, mostly in the vicinity of Langara Island, during May-July. Eleven stock regions are represented in these samples, from the Skeena River in the north to Coastal Oregon in the south. The South Thompson region was the most common with 38% of samples, but the regions within the Columbia River system were also important, totalling 24% of samples. A significant portion of Chinook were from regions in closer proximity, such as the North Mainland (18%) and three regions in the Skeena River system (total of 9%).

North mainland coast – PFMA 3-6

A total of 52 Chinook salmon kills sampled during May-July in the north mainland coast area (Figure 10) were assigned to stock regions. Sixteen stock regions were represented, with the West Coast Vancouver Island and South Thompson regions being most important. Although distant regions such as the Upper Columbia River and Coastal Oregon were also represented, a greater proportion (38%) of Chinook originated in local regions in the Skeena River system, Nass, and North Mainland.

Central coast area – PFMA 7-11

Only 24 Chinook samples were available for this area, as shown in Figure 11. Of these, almost half (n = 11) were Chinook from local stocks in the North Mainland region. Seven fish were from the Columbia River system; these were all collected in outer coast waters off the western entrance to Queen Charlotte Strait.

6 377

Northeastern Vancouver Island – PFMA 12-13

The greatest number of Chinook salmon samples (n = 205) were collected from this area, which includes the waters of eastern Queen Charlotte Strait and Johnstone Strait, which have been designated as Critical Habitat for northern resident killer whales (Fisheries and Oceans Canada 2008). Most samples were collected during July-September. Well over half (62%) of Chinook salmon sampled in this area were from stocks within the Fraser River system (Figure 12). Of these, the South Thompson region was the particularly important with 42% of overall samples. Other significant Chinook regions were East Coast Vancouver Island (16%) and West Coast Vancouver Island (8%).

Southeastern Vancouver Island – PFMA 14-19, 28-29

Despite the small sample size for this region (n = 24), they were collected over 8 months of the year, May-November and January (Figure 13). Some of these samples were collected in U.S. waters near San Juan Island and in Puget Sound. As would be expected, all Chinook sampled in this inshore area were from local regions, mostly the Fraser River system (n = 16) but also the East Coast Vancouver Island, South Mainland and Puget Sound regions. Stock regions identified in this area are consistent with those identified from more extensive Chinook samples collected from southern residents by Hanson et al. (in press).

West coast Vancouver Island – PFMA 20-27

A total of 136 Chinook samples were collected in this area, mostly in or near the entrance to Juan de Fuca Strait (Figure 14). Most samples were collected within waters designated as Critical Habitat for southern resident killer whales in Canadian waters. More than three-quarters of these Chinook were from stocks within the Fraser River system, with the South Thompson being clearly the most important region (39% of the total samples). Other regions represented include Puget Sound (13%) and West Coast Vancouver Island (7%).

EFFECTS OF VARIATIONS IN CHINOOK SALMON AVAILABILITY

Chinook salmon plays such an important role in the diet of resident killer whales that this prey species’ availability appears to affect the whales’ survival and, to a lesser extent, reproductive rates. Ford et al. (2005, 2010) showed that over the period of 1979-2003, there was a strong negative correlation between the mortality rates of both northern and southern resident killer whales and the coast-wide abundance of Chinook salmon, and a lesser though still significant positive correlation with calving rates. Particularly striking was the period of very low Chinook abundance in the mid to late 1990s, which was correlated with mortality rates 2-3 times higher than expected in resident killer whales. Unusually high mortalities were observed broadly among different killer whale groups and different age/sex categories.

The longer time series of data now available on resident killer whale population dynamics and foraging behaviour has allowed us to further investigate the effects of fluctuations in Chinook salmon abundance.

7 378

Mortality rates versus Chinook salmon abundance

Comparisons between the population dynamics of resident killer whales and Chinook salmon abundance described in Ford et al. (2005, 2010) extended from 1979, the first year of the annual Pacific Salmon Commission (PSC) Chinook abundance, to 2003, the last year for which whale population data were available. With an additional five years of data now available, we have repeated the correlation analysis between mortality and Chinook abundance to determine whether the same relationship has been maintained in recent years. The procedure used for this analysis is essentially the same as that described in Ford et al. (2005). An index of mortality was derived by calculating the ratio of the number of deaths observed in the populations to the number expected for each year. The number of expected mortalities was calculated from sex- and age-specific mortality schedules provided in Olesiuk et al. (2005), for a period of unrestrained growth during 1973-96. Because there was sometimes uncertainty associated with exact year of death of some individuals, and deaths might be influenced by effects that were cumulative over several years (e.g. nutritional stress), we expressed the observed to expected ratios as 3-year running averages. These annual mortality indices were compared to an annual index based on the total estimated Chinook salmon abundance across six coastal regions, developed by the PSC Chinook Technical Committee (PSC 2008). In our earlier analyses (Ford et al. 2005, 2010), we modified the PSC Chinook abundance index by referencing each year’s abundance to the average annual abundance over the 1979-2003 time series, rather than the 4-year base period (1979-82) used by the PSC. In the current analysis, we have used the PSC index without modification.

The mortality indices for northern and southern resident killer whales and annual PSC Chinook abundance indices for the 1979-2008 period are depicted in Figure 15. An updated regression analysis for this time series (Figure 16), with killer whale mortalities lagged by one year following salmon abundance, confirms the strong relationship between killer whale survival and coast-wide Chinook abundance. Although the correlation is not as strong as for the earlier 2 2 period (r = 0.777 (1979-2003) vs r = 0.487 (1979-2008)), it is still highly significant (F1,27 = 25.6, p < 0.001). In the years since the earlier analysis, the Chinook salmon abundance index has fluctuated considerably, from values above the base period in 2003-04, to below the base period in 2005-08. On average over the 1979-2008 period, a killer whale mortility index above 1 was associated with a coast-wide Chinook abundance index of 1.1 or less (calculated from regression in Figure 16). Despite lower Chinook abundance in the most recent years, killer whale mortality rates have yet to exhibit an increase.

Prey selection versus Chinook salmon abundance

During the period of unusually low Chinook salmon abundance and high resident killer whale mortality rates in the mid to late 1990s, there was very little field effort to document foraging behaviour and diet (Figure 3). As a result, there is no information available on the effect this reduced abundance of the whales’ preferred prey may have had on prey selection or foraging behaviour. Since 2002, however, substantial effort has been dedicated to such studies and during this period the abundance of Chinook has fluctuated considerably. Table 5 presents salmonid prey species taken during foraging in each of these years, together with the total PSC index for that year. Despite a Chinook salmon abundance as high as 1.3 in 2003, and as low as 0.62 in 2007, no shift in prey species composition is evident. No significant difference was found when comparing the proportions of Chinook salmon to chum salmon taken in years of high Chinook abundance (2003-06) versus low Chinook abundance (2007-08) (Fisher’s exact test, p = 0.16). Coho salmon comprised 10.5% of total kills in 2008, but 7 of the 10 samples were from a single day and there was no broad shift to this species.

8 379

ESTIMATED CONSUMPTION RATES OF CHINOOK SALMON

To assess the quantity of Chinook salmon consumed by resident killer whales, our analysis involved the following four general steps:

1) Estimate the metabolic needs of resident killer whales 2) Estimate the caloric value of Chinook salmon consumed by killer whales based on prey size and energy density 3) Estimate the proportion of the whales’ diet that is composed of Chinook salmon 4) Estimate the total numbers of Chinook salmon consumed by the current northern and southern resident killer whale populations

Metabolic needs of resident killer whales

Estimation of the energetic requirements of killer whales requires accurate estimates of the body mass of individuals based on age and sex. Our methods generally follow those of Noren (in press), who has recently undertaken a similar assessment of the energetic requirements of southern resident killer whales. Whales less than 1 year old were discounted from these analyses as they were assumed to be completely dependent on their mother for nourishment. For whales 1-12 years old, sexual dimorphism is minimal so males and females were combined in the same categories based on age in years. For whales 13 years or older, separate categories were made for each sex by age in years to account for sexual dimorphism and thus the greater energetic requirements of adolescent and adult males as compared to females of the same age. Changes in energetic requirements for adult females during pregnancy and lactation were not accounted for in the calculations.

Estimated body mass values were then determined for each age and sex class. For whales aged ≤12 years, body mass was calculated (as in Noren, in press) using a formula that estimates body weight based on age in days for female killer whales in aquaria (Clark et al. 2000):

(1) where y = body mass in kg, x = age in days, and exp = e raised to the power of a given number. In Noren (in press), body masses for southern resident whales ≥13 years were estimated using maximum lengths of whales measured during the live-capture fishery in British Columbia and Washington state from 1962-1973 (Bigg and Wolman 1975). In order to estimate terminal adult body masses (≥20 years) for resident killer whales in this analysis, we used the average lengths for southern resident males (677 cm) and females (600 cm) 20 years of age or older, measured by Durban et al. (2009) using aerial photogrammetric techniques. Although measurements of killer whales are available from other sources, we chose these measurements as being the least biased and most applicable since they were taken from the same (in the case of southern residents) or a closely related (in the case of northern residents) population to the whales in this study. Body lengths were next converted to mass as in Noren (in press), using an equation developed by Bigg & Wolman (1975) from measurements of live-captured killer whales:

(2) where M = body mass in kg, and L = length in cm. Once terminal body masses for adult whales ≥20 years had been determined (Table 6), the estimated body mass for each of the intervening

9 380

age and sex-classes was calculated, assuming a constant yearly growth rate between the estimated mass at 12 years (calculated from equation of Clark et al. 2000) until terminal body mass was reached at 20 years (Table 6).

The daily prey energetic requirements (DPERs) of individual resident killer whales were calculated using formulae developed by Noren (in press):

(3) (4)

where DPER = daily prey energy requirements in kcal/d and Mb = body mass in kg. Minimum and maximum values of DPER reflect the range of field metabolic rates estimated by Noren (in press) to be 5 to 6 times the basal metabolic rates predicted for mammals by Kleiber (1975). DPER values also take into account digestive efficiency for killer whales, which is estimated to be about 84.7% (Williams et al. 2004). Killer whales must therefore consume an additional 15.3% of their estimated field metabolic rate value each day in order to meet their energy requirements.

Results of DPER calculations for individuals by age- and sex-class are presented in Table 6 and Figure 17. The DPER values for individual whales were multiplied by the number of whales in each age/sex-class to obtain DPER values for the entire population. These calculations were done separately for northern (Table 7, Figure 18) and southern (Table 8, Figure 19) residents to account for differences in population size and demography. The resulting range of DPER values for the entire northern resident population (n = 241 animals ≥ 1 year old, 2008 census) is 34,025,721 - 40,835,806 kcal/day. DPER for the southern resident population (n = 85 animals ≥ 1 year old, 2008 census) is 12,753,120 - 15,305,596 kcal/day.

Energetic value of Chinook salmon consumed by resident killer whales

In order to convert DPER of resident killer whales into the number of Chinook salmon required to sustain each of the two populations, we first needed to determine the energetic value of Chinook salmon consumed by the whales. Because a range of age classes (and therefore sizes) of Chinook salmon is taken by resident killer whales, the caloric content of each prey item may vary widely. A profile of ages determined for Chinook killed by northern and southern residents is presented in Table 9. Energy content of different age classes of Chinook salmon was determined using average fork lengths by age from Ford and Ellis (2006) and a regression of fork length to energy content developed by O’Neill et al. (in prep.; Figure 20). By dividing DPER values by the estimated energy content per fish for each age class of Chinook salmon, we calculated the number of fish each of age class that each resident population would have to consume in order to meet their daily energy requirements, assuming a diet of 100% Chinook salmon. The results for northern residents are shown in Table 11 and southern residents in Table 12.

Estimated numbers of Chinook salmon consumed by resident killer whales

Assuming a 100% Chinook salmon diet, the estimated daily requirement for northern resident killer whales is 3063 - 3676 Chinook salmon, and for southern residents, 1338 - 1606 Chinook salmon. Our estimates for southern residents are higher than the 775 - 928 Chinook per day calculated by Noren (in press) for this population, likely as a result of our incorporating

10 381

the age structure of Chinook salmon actually taken by killer whales into our estimates, rather than using an average mass and caloric value for adult fish.

Although prey sampling suggests that Chinook salmon may well represent 100% of the whales’ diet at certain locations and times of the year, this is clearly not the case on an annual basis. Chinook is the predominant prey species observed in most regions (Figure 8) and during most months (Figure 7), but chum salmon are more important than Chinook during their spawning migration in October. Diet in winter is also poorly known, although Chinook salmon still appear to be targeted by resident whales from the few feeding events sampled. Although non-salmonids such as Pacific halibut, lingcod, and Dover sole (Microstomus pacificus) appear to comprise a relatively small component of the diet of resident killer whales based on prey sampling and stomach contents analysis (Ford et al. 1998; Ford and Ellis 2005; Hanson et al. in press), such demersal prey may be more important during winter. As a result, estimates of the annual consumption rate of Chinook salmon by resident killer whales are fraught with uncertainty.

Despite these uncertainties, it is useful to provide some estimates of the potential range of annual consumption rates of Chinook salmon by resident killer whales to assess its magnitude with respect to coast-wide Chinook abundance and harvest levels in fisheries. Tables 13 and 14 present estimates of the daily and annual Chinook salmon consumption by northern and southern resident killer whales, respectively, at levels between 50 and 100% diet composition. At the 70% level, which corresponds to the overall proportion of Chinook salmon in prey samples identified to species (Table 2), northern residents would require 782,482 to 939,092 Chinook per year and southern residents 341,917 to 410,350 per year. At this predation level, the two populations at their current abundance would consume a total of about 1,124,000 to 1,350,000 Chinook salmon annually.

Key foraging habitats for resident killer whales are the waters of eastern Queen Charlotte Strait and Johnstone Strait (northern residents) and the waters of the Strait of Georgia, Juan de Fuca Strait, and Puget Sound (southern residents) (Ford et al. 2000; Krahn et al. 2004). These areas have been designated as Critical Habitats (CH) under the Species-at- Risk Act in Canada (for waters in Canada’s jurisdiction) and the Endangered Species Act in the U.S. (for U.S. waters) (Ford 2006; NMFS 2006; Fisheries and Oceans Canada 2008). The whales use these CH areas predominantly during summer and fall, and their occurrence coincides with that of migrating Chinook salmon (Ford and Ellis 2006; Hanson et al. in press). Both of these areas are also used extensively for commercial and recreational salmon fishing, the latter of which generally targets Chinook salmon. We assessed the probable numbers of Chinook salmon taken by killer whales in CH during July-August, as these are the peak months of resident killer whale occurrence and their diet is predominantly Chinook salmon during this period (87% of prey samples in northern resident CH (this study) and 91% of prey samples in southern resident CH (Hanson et al. in press).

To estimate the quantity of Chinook salmon preyed upon by resident killer whales in their Critical Habitats, we calculated the number of days that whales are typically present in the two CH areas during July-August, and then multiplied this by the DPER values provided in Table 7 according to the numbers and age- and sex-class of killer whales present on those days. Total DPER of whales each day was then converted to number of Chinook salmon as in Tables 11 and 12. It was assumed that 90% of the whales’ diet was composed of Chinook salmon, which corresponds closely to the proportion of Chinook in identified prey samples during this period for northern residents (87%, this study) and southern residents (91%, Hanson et al. in press). It should be noted that Chinook may represent more than 90% of the whales’ energetic

11 382

requirements during this period, as Chinook salmon are larger and have higher energy densities per fish than do fish making up the remaining 10% of prey species (coho, pink, sockeye, and chum salmon).

The entire southern resident population tends to spend the July-August period within Canadian and US waters designated as CH in their respective jurisdictions. Hauser et al. (2007) noted that the population can be found in inshore waters of eastern Juan de Fuca Strait, the Strait of Georgia and Puget Sound on 80% of summer days. It is probable that on the majority of the remaining 20% of days, the whales mostly use western Juan de Fuca Strait, which is within the CH boundaries, or areas on Swiftsure Bank, which is near the entrance to Juan de Fuca Strait but outside CH (Ford 2006; Cetacean Research Program, Pacific Biological Station, unpubl. data). We thus assumed that the southern resident population is present in CH waters on a minimum of 90% of days in July-August. The resulting estimated Chinook salmon requirement for southern residents in their CH (in both U.S. and Canadian waters) during July- August is 1204 to 1445 fish per day, or approximately 67,000 to 81,000 fish over the two month period.

Estimating Chinook salmon requirements of northern residents in their CH was not as straightforward as for southern residents because only a portion of the population utilizes these waters, even during the peak months of July-August. As a result, we calculated an average total number of ‘whale days’ (number of whales present in CH per day) for the July-August period based on whale occurrence during these months in 1998-2008. An average of 32.1 whales (± 2.9 SE) were present in CH per day during July-August, 1998-2008, which represents 14.5% of the average 222 animals in the population across those years. These ‘whale days’ were then partitioned according to the average demographic composition for resident killer whales (Olesiuk et al. 2005), the mean DPER for these age- and sex-classes was applied, and the number of Chinook salmon required was calculated as described above. This analysis resulted in an estimated Chinook salmon requirement for northern residents in their CH during July-August of 419-503 fish per day, or approximately 26,000 to 31,000 fish over the two month period.

The estimates of total annual Chinook salmon predation provided in Tables 13 and 14 reflect the requirements of northern and southern resident killer whales in 2008. As continued population growth is considered a priority in the Recovery Strategy for Resident Killer Whales, we have estimated the potential future requirements for Chinook salmon by the two populations assuming optimal growth over the next decade. For this analysis, we assumed an annual growth of 2.6%, the rate observed in the populations during the period of unrestrained growth between 1973 and 1995, and an average age- and sex-class composition described by Olesiuk et al. (2005). The annual increase in Chinook salmon requirements, assuming a 70% diet composition, over the period 2008-2018, is shown in Figure 21. By 2018, the total abundance of resident killer whales would be 445 (332 northern and 113 southern residents), and their annual requirement for Chinook salmon would be in the range of 1,480,000 to 1,780,000 fish.

DISCUSSION

Since our last assessment of resident killer whale diet (Ford and Ellis 2005, 2006), we have doubled the dataset on salmonids identified from feeding events, from 396 to 806. Larger sample sizes are now available for most coastal regions, especially for important feeding areas for southern residents off southwestern Vancouver Island. Although few in number, we now have prey samples collected from resident feeding events during the winter months. This much larger dataset provides further support for our past conclusions: that Chinook salmon is clearly

12 383

the preferred and most important prey of resident killer whales and that the smaller pink and sockeye salmon are not significant prey despite their greater seasonal abundance. Chum salmon is also an important species, particularly during the fall as these fish migrate through inshore waters. Coho salmon make up a small portion of the whales’ catches in some regions, but only represent 2.5% of prey items overall. Of the 7 feeding events sampled in December- March, 5 were Chinook salmon, 1 was a chum salmon, and 1 was a steelhead salmon.

Chinook salmon are not only numerically the most frequently consumed prey species in most months, they are also generally larger in body than other salmonids. Killer whales appear to select for large Chinook, and so most prey are 4-5 year old fish which have mean body masses of 8-13 kg (Ford and Ellis 2005). This is considerably larger than mature chum salmon (4.0-5.5 kg) and more than double the typical size of coho, pink and sockeye salmon (Ford et al. 1998). As Chinook salmon also tend to have the highest lipid content of salmonids, the energy content per fish is considerably greater than other salmonid species.

The updated assessment of the relationship between mortality rates of resident killer whales and coast-wide Chinook salmon abundance has continued to show that there is a significant long-term negative correlation. The correlation is not as strong in recent years, due primarily to low mortality and continued growth in the northern resident population despite relatively low Chinook abundance since 2006 (Pacific Salmon Commission 2008). On-going annual monitoring will determine whether this trend continues.

Our observations of foraging behaviour of resident killer whales suggest that fluctuations in coast-wide Chinook salmon availability have little effect on prey selection, at least during the summer months. Further studies are needed to quantitatively assess the balance between the energetic costs of foraging in low Chinook density conditions and the caloric value of prey obtained.

Genetic stock identification of Chinook salmon prey samples indicates that resident killer whales consume fish originating from a wide variety of coastal regions, some quite distant from the place of capture. For example, almost one-quarter of Chinook taken by residents off the northern Queen Charlotte Islands originated from the Columbia River, the mouth of which is over 1000 km to the south. This is consistent with northern British Columbia troll catches where the predominant stock groupings were South Thompson, followed by North and Central Oregon and Upper Columbia Summer and Fall (Winther and Beacham in press). Stocks from the Fraser River system were represented most frequently in feeding events in most parts of the coast, and comprised 58% of samples overall. This is not unexpected, given that the Fraser River system is the largest producer of Chinook salmon in Canada (Parken et al. 2008). The predominance of Fraser River Chinook was particularly notable in samples collected from feeding events in Critical Habitat areas off northeastern and southwestern Vancouver Island. Fraser River stock regions comprised 64% of Chinook consumed by northern residents in their Critical Habitat, and 75% of Chinook taken by southern residents in their Critical Habitat. South Thompson was the most prevalent of the Fraser River stock regions.

Although the Fraser River system may be the most important source of Chinook salmon for resident killer whales generally, other stock regions may also be important at certain times of year. Chinook originating from smaller, local river systems were significant prey of resident killer whales along the north and central mainland coasts. Many of the northern resident groups feed in these areas early in the summer, before moving south to the Critical Habitat area off northeastern Vancouver Island later in the summer (Ford 2006). Thus, whales may rely on a range of Chinook stocks at different times of year and in different parts of the coast. The results

13 384

presented here are preliminary; further effort is needed to determine the seasonal importance of particular Chinook salmon stocks to whales in different geographic areas, and the conservation status of these stocks should be evaluated in this context.

Our assessment of the quantity of Chinook salmon needed to sustain current resident killer whale populations indicates that a substantial number of fish may be consumed each year. Although there is considerable uncertainty in the actual proportion of the whales’ year-round diet that is composed of Chinook, a reasonable conservative estimate is that about 70% of their nutritional needs may be supplied by this species. If this is the case, consumption by the current resident populations may be over 1,000,000 fish per annum (range of estimate 1,124,400 to 1,349,443). This is roughly equivalent to the total combined commercial and recreational harvest of Chinook salmon in marine waters between Southeast Alaska and Oregon during 2006 (Pacific Salmon Commission 2007; R. McNicol, Pacific Biological Station, pers. comm.).

If resident killer whales are near the carrying capacity of their habitat and if that capacity is determined by the availability of Chinook salmon, as the correlation between mortality and Chinook abundance implies (Ford et al. 2005), then greater numbers of Chinook salmon will be required to provide for recovery. Assuming that resident killer whale populations grow at their maximum rate of 2.6% over the next 10 years, an estimated 1.5-1.8 million Chinook may be needed to support these populations each year by 2018.

Critical Habitats that have been designated for northern and southern resident whales under Canada’s Species at Risk Act and the U.S. Endangered Species Act are prime feeding areas during the peak of the summer salmon migration period. It is thus imperative that sufficient prey resources be available to the whales in these areas at this important time of year. Extensive prey sampling in Critical Habitats suggests that Chinook salmon represents about 90% of resident killer whale diet during July-August. Southern residents foraging in Critical Habitat (in Canadian and U.S. waters combined) would thus require approximately 1200 - 1400 Chinook salmon per day, or roughly 67,000 - 81,000 over the two month period. On average, only 14.5% of the northern resident population uses their designated Critical Habitat on a daily basis during July-August. As a result, Chinook salmon requirements in this area are less than for southern resident Critical Habitat: about 420 - 500 fish per day, or 26,000 - 31,000 total over the two months. As the great majority of Chinook taken in both Critical Habitat areas are from Fraser River stocks, it can be concluded that adequate Chinook production in this river system is essential to the continued function of resident killer whale Critical Habitats.

It should also be noted that estimates of Chinook salmon consumption rates are based on the whales’ predicted daily prey energy requirements. However, it may well be that during certain times of year, especially during the summer Chinook salmon migration, the whales feed at rates that surpass their daily requirements and, in so doing, create blubber reserves that are needed during periods of reduced prey availability. Further research on foraging behaviour and prey selection, particularly in winter and spring, is necessary to better understand the year- round prey composition, feeding rates, and energetics of resident killer whales.

14 385

ACKNOWLEDGEMENTS

We thank the many research colleagues and volunteers who helped in the collection of data reported here, including R. Abernethy, E. Ashe, K. Balcomb, L. Barrett-Lennard, J. Borrowman, L. Dalla Rosa, V. Deecke, M., B., & R. De Roos, J. Durban, B. Falconer, B. Gisborne, J. Hildering, M. Malleson, C. McMillan, L. Nichol, B. Paterson, L. Spaven, J. Towers, J. Watson, R. Williams, and H. Yurk. K. Balcomb, Center for Whale Research, provided updated demographic data for southern resident killer whales. S. MacLellan and staff in the Sclerochronology Lab at the Pacific Biological Station (PBS) undertook species identification and ageing, and the laboratory staff at the Molecular Genetics Lab at PBS conducted the DNA analyses. D. Noren, Northwest Fisheries Science Center, NOAA, kindly allowed us to refer to her manuscript on energetics of southern resident killer whales. R. Abernethy prepared the maps and M. Boogaards helped with data entry. We appreciate the assistance of G. Brown, R. McNicol, and C. Parken in providing data and information on Chinook salmon.

LITERATURE CITED

Altmann, J. 1974. Observational study of behavior: sampling methods. Behaviour 49: 227-265.

Barrett-Lennard, L.G. 2000. Population structure and mating patterns of killer whales (Orcinus orca) as revealed by DNA analysis. Ph.D. Thesis, University of British Columbia, Vancouver, BC. xi + 97 p.

Beacham, T.D., Candy, J.R., Supernault, K.J., Wetklo, M., Deagle, B., Labaree, K., Irvine, J.R., Miller, K.M., Nelson, R.J., and Withler, R.E. 2003. Evaluation and application of microsatellites for population identification of Fraser River chinook salmon (Oncorhynchus tshawytscha). Fish. Bull. 101: 243-259.

Beacham, T.D., Candy, J.R., Jonsen, K.L., Supernault, J., Wetklo, M., Deng, L., Miller, K.M., Withler, R.E., and Varnavskaya, N. 2006. Estimation of stock composition and individual identification of Chinook salmon across the Pacific Rim using microsatellite variation. Trans. Amer. Fish. Soc. 135: 861-888.

Bigg, M.A. 1982. An assessment of killer whale (Orcinus orca) stocks off Vancouver Island, British Columbia. Rep. Int. Whal. Commn. 32: 655-666.

Bigg, M.A., Ellis, G.M., Ford, J.K.B., and Balcomb, K.C. 1987. Killer Whales: A study of their identification, genealogy and natural history in British Columbia and Washington State. Phantom Press and Publishers, Nanaimo, BC. 79 p.

Bigg, M.A., MacAskie, I.B., and Ellis, G. 1976. Abundance and movements of killer whales off eastern and southern Vancouver Island with comments on management. Unpubl. Rep., Arctic Biological Station, Dept. of Fisheries and Environment, Ste Anne-de-Bellevue, QC.

Bigg, M.A., Olesiuk, P.F., Ellis, G.M., Ford, J.K.B., and Balcomb, K.C. III. 1990. Social organization and genealogy of resident killer whales (Orcinus orca) in the coastal waters of British Columbia and Washington State. Rep. Int. Whal. Commn. Special Issue 12: 383-405.

15 386

Bigg, M.A., and Wolman, A.A. 1975. Live-capture killer whale (Orcinus orca) fishery, British Columbia and Washington, 1962-73. J. Fish. Res. Bd Can. 32: 1213-1221.

Clark, S.T., Odell, D.K., and Lacinak, C.T. 2000. Aspects of growth in captive killer whales (Orcinus orca). Mar. Mamm. Sci. 16: 110-123.

Dahlheim, M.E., Schulman-Janiger, A., Black, N., Ternullo, R., Ellifrit, D., and Balcomb, K.C. III. 2008. Eastern temperate North Paciﬁc offshore killer whales (Orcinus orca): occurrence, movements, and insights into feeding ecology. Mar. Mamm. Sci. 24: 719-729.

Durban, J., Fearnbach, H., Ellifrit, D. and Balcomb, K. 2009. Size and body condition of southern resident killer whales. Contract report to the Northwest Regional Office, National Marine Fisheries Service. Order number AB133F08SE4742, Requisition number NFFP5000-8-43300. 21 p.

Fisheries and Oceans Canada. 2008. Recovery Strategy for the Northern and Southern Resident Killer Whales (Orcinus orca) in Canada. Species at Risk Act Recovery Strategy Series, Fisheries & Oceans Canada, Ottawa, ON. ix + 80 p.

Ford, J.K.B. 1989. Acoustic behaviour of resident killer whales (Orcinus orca) off Vancouver Island, British Columbia. Can. J. Zool. 67: 727-745.

Ford, J.K.B. 1991. Vocal traditions among resident killer whales (Orcinus orca) in coastal waters of British Columbia. Can. J. Zool. 69: 1454-1483.

Ford, J.K.B. 2006. An assessment of critical habitats of resident killer whales in waters off the Pacific coast of Canada. DFO Can. Sci. Advis. Sec. Res. Doc. 2006/072. iv + 34 p.

Ford, J.K.B., and Ellis, G.M. 2005. Prey selection and food sharing by fish-eating ‘resident’ killer whales (Orcinus orca) in British Columbia. DFO Can. Sci. Advis. Sec. Res. Doc. 2005/041. ii + 30 p.

Ford, J.K.B., and Ellis, G.M. 2006. Selective foraging by fish-eating killer whales Orcinus orca in British Columbia. Mar. Ecol. Prog. Ser. 316: 185-199.

Ford, J.K.B., Ellis, G.M. and Balcomb, K.C. 2000. Killer Whales: The natural history and genealogy of Orcinus orca in British Columbia and Washington, Second edition. UBC Press and University of Washington Press, Vancouver, BC and Seattle, WA. 104 p.

Ford, J.K.B., Ellis, G.M., Barrett-Lennard, L.G., Morton, A.B., Palm, R.S., and Balcomb, K.C. III. 1998. Dietary specialization in two sympatric populations of killer whales (Orcinus orca) in coastal British Columbia and adjacent waters. Can. J. Zool. 76: 1456-1471.

Ford, J.K.B., Ellis, G.M., and Olesiuk, P.F. 2005. Linking prey and population dynamics: Did food limitation cause recent declined of ‘resident’ killer whales (Orcinus orca) in British Columbia? DFO Can. Sci. Advis. Sec. Res. Doc. 2005/042. ii + 27 p.

Ford, J.K.B, Ellis, G.M., Olesiuk, P.F., and Balcomb, K.C. 2010. Linking killer whale survival and prey abundance: food limitation in the oceans’ apex predator? Biol. Lett. 6:139-142. Published on-line before print September 15, 2009.

16 387

Guinet, C. 1990. Sympatrie des deux catégories d'orques dans le détroit de Johnstone, Colombie Britannique. Rev. Ecol. (Terre Vie) 45(1): 25-34.

Hanson, M.B., Baird, R.W., Ford, J.K.B., Hempelmann-Halos, J., Van Doornik, D., Candy, J.R., Emmons, C.K., Schorr, G.S., Gisborne, B., Ayres, K.L., Wasser, S.K., Balcomb, K.C., Balcomb-Bartok, K., Sneva, J.G., and Ford, M.J. In press. Species and stock identification of prey selected by endangered “southern resident” killer whales in their summer range. End. Species Res.

Hauser, D.D.W., Logsdon, M.G., Holmes, E.E., VanBlaricom, G.R., and Osborne, R.W. 2007. Summer distribution patterns of southern resident killer whales Orcinus orca: core areas and spatial segregation of social groups. Mar. Ecol. Prog. Ser. 351: 301-310.

Healey, M.C. 1991. Life history of chinook salmon (Oncorhynchus tshawytscha). In Pacific Salmon Life Histories. Edited by C. Groot and L. Margolis. UBC Press, Vancouver, BC. pp. 313-393.

Heimlich-Boran, J. 1986. Fishery correlations with the occurrence of killer whales in greater Puget Sound. In Behavioral Biology of Killer Whales. Edited by B.C. Kirkevold and J.S. Lockard. Alan R. Liss, Inc., New York, NY. pp. 113-131.

Kleiber, M. 1975. The Fire of Life: An Introduction to Animal Energetics, Second edition. Robert E. Kreiger Publishing, Huntington, NY. 478 p.

Krahn, M.M., Ford, M.J., Perrin, W.F., Wade, P.R, Angliss, R.P., Hanson, M.B., Taylor, B.L., Ylitalo, G.M., Dahlheim, M.E., Stein, J.E., and Waples, R.S. 2004. 2004 status review of southern resident killer whales (Orcinus orca) under the Endangered Species Act. U.S. Dept. of Commerce, NOAA Tech. Memo. NMFSNWFSC-62. xviii + 73 p.

MacLellan, S.E. 2004. Guide for sampling structures used in age determination of Pacific salmon. Pacific Biological Station, Fisheries and Oceans Canada, Nanaimo, British Columbia V9T 6N7.

Mann, J. 1999. Behavioral sampling methods for cetaceans: a review and critique. Mar. Mamm. Sci. 15(1): 102-122.

Neaves, P.I., Wallace, C.G., Candy, J.R., and Beacham. T.D. 2005. CBayes: Computer program for mixed stock analysis of allelic data. Version v5.01. Free program distributed by the authors over the internet from http://www.pac.dfo- mpo.gc.ca/sci/mgl/Cbayes_e.htm

Nichol, L.M. and Shackleton, D.M. 1996. Seasonal movements and foraging behaviour of northern resident killer whales (Orcinus orca) in relation to the inshore distribution of salmon (Oncorhynchus spp.) in British Columbia. Can. J. Zool. 74: 983-991.

NMFS. 2005. Endangered and Threatened Wildlife and Plants: Endangered Status for Southern Resident Killer Whales, Federal Register, 70 FR 69903.

NMFS. 2006. Designation of critical habitat for southern resident killer whales: Biological Report, October 2006. 44 p. Available at http://www.nwr.noaa.gov/Marine-Mammals/Whales- Dolphins-Porpoise/Killer-Whales/ESA-Status/upload/SRKW-CH-Bio-Rpt.pdf

17 388

Noren, D.P. In press. Estimated field metabolic rates and prey requirements of resident killer whales. Mar. Mamm. Sci.

Olesiuk, P.F., Bigg, M.A., and Ellis, G.M. 1990. Life history and population dynamics of resident killer whales (Orcinus orca) in the coastal waters of British Columbia and Washington State. Rep. Int. Whal. Commn. Special Issue 12: 209-242.

Olesiuk, P.F., Ellis, G.M. and Ford, J.K.B. 2005. Life history and population dynamics of northern resident killer whales (Orcinus orca) in British Columbia. DFO Can. Sci. Advis. Sec. Res. Doc. 2005/045. iv + 75 p.

O’Neill, S.M., Ylitalo, G.M., West, J.E., Bolton, J., Sloan, C.A. and Krahn, M.M. In prep. Regional patterns of persistent organic pollutants in five Pacific salmon species (Oncorhynchus spp.) and their contributions to contaminant levels in northern and southern resident killer whales (Orcinus orca). NOAA, National Marine Fisheries Service, Northwest Fisheries Science Center.

Pacific Salmon Commission. 2007. 2007 annual report of catches and escapements, exploitation rate analysis and model calibration. Joint Chinook Technical Committee Report TCCHINOOK (07)-2.

Pacific Salmon Commission. 2008. 2008 annual report of catches and escapements, exploitation rate analysis and model calibration. Joint Chinook Technical Committee Report TCCHINOOK (08)-2.

Parken, C.K., Candy, J.R., Irvine, J.R., and Beacham, T.D. 2008. Genetic and coded wire tag results combine to allow more-precise management of a complex Chinook salmon aggregate. N. Am. J. Fish. Manage. 28(1): 328-340.

Ward, E.J., Holmes, E.E., and Balcomb, K.C. 2009. Quantifying the effects of prey abundance on killer whale reproduction. J. Appl. Ecol. 46(3): 632-640.

Wiles, G. J. 2004. Washington State status report for the killer whales. Washington Department of Fish and Wildlife, Olympia, WA. 106 p.

Williams, T.M., Estes, J.A., Doak, D.F., and Springer, A.M. 2004. Killer appetites: assessing the role of predators in ecological communities. Ecology 85(12): 3373-3384.

Winther, I., and Beacham, T.D. In press. The application of Chinook salmon stock composition data to management of the Queen Charlotte Islands troll fishery, 2006. In C. C. Krueger, and C. E. Zimmerman, editors. Sustainability of the Arctic-Yukon-Kuskokwim salmon fisheries: what do we know about salmon ecology, management, and fisheries? American Fisheries Society, Symposium XX, Bethesda, Maryland.

18 389

TABLES

Table 1. Evidence for 937 feeding events by resident killer whales documented during 1974-2009.

Evidence of predation Number of events %

Observation only 78 8.3 Both tissue and scale 28.1 263 samples Tissue samples only 54 5.8 Scale samples only 542 57.9 Total 937 100

Table 2. Species composition of fish killed by resident killer whales in 937 feeding events during 1974- 2009 in different B.C. coastal regions. Species identity was determined by scale analysis or from DNA evidence. Also shown is the frequency distribution (%) of different salmonid species among the 806 salmonid prey items identified to species. PFMA refers to the Pacific Fisheries Management Areas of Fisheries & Oceans Canada. UnSa are salmonids that were observed as prey in the field but not sampled for species identification, or salmonids that could not be identified to species. UnFi are fish that could not be positively identified to species and could include either salmonids or non-salmonids.

Region PFMA n Species Chinook Chum Coho Pink Sockeye Steelhead Other UnSa UnFi

QC Island 1-2 40 36 0 0 0 0 0 2a 5 3

North coast 3-6 80 54 20 0 0 0 0 0 5 1

Central coast 7-11 72 50 6 1 1 0 0 0 11 3

NE Vancouver Island 12-13 507 247 159 7 13 2 0 6b 62 10

SE Vancouver 14-19, Island 28-29 61 36 5 2 0 1 3 1c 12 3

W Vancouver Island 20-27 171 146 5 10 0 1 1 1d 2 4

Total 937 569 195 20 14 4 4 10 97 24 % of identified salmonids 71.0 23.8 2.5 1.7 0.5 0.5 a – 1 Pacific halibut, 1 herring b – 1 yelloweye rockfish, 4 sardine, 1 herring c – 1 quillback rockfish d – 1 sablefish

19 390

Table 3. Salmonid species sampled from resident killer whale feeding events by month, 1974-2009. n = 806 feeding events.

Month Species Total Chinook Chum Coho Pink Sockeye Steelhead Jan 4 0 0 0 0 0 4 Feb 1 0 0 0 0 0 1 Mar 1 0 0 0 0 0 1 Apr 0 0 0 0 0 0 0 May 21 0 0 0 0 0 21 Jun 96 15 0 0 0 0 111 Jul 159 16 1 1 3 1 181 Aug 212 4 5 11 1 0 233 Sep 60 59 10 2 0 1 133 Oct 12 96 4 0 0 1 112 Nov 3 4 0 0 0 0 7 Dec 0 1 0 0 0 1 2 Total 569 195 20 14 4 4 806

Table 4. Ages of 634 salmonids (identified to species level) determined from scales collected from feeding events by resident killer whales during 1974-2009. Age classes are given according to the European system: the years spent in freshwater after hatching preceed the years in salt water, separated by a decimal point.

European Age Class Species n 0.1 0.2 0.3 0.4 0.5 1.1 1.2 1.3 1.4 1.5 2.1 2.2 Chinook 431 1 41 179 101 3 2 26 55 20 2 0 1 Chum 180 0 2 122 52 4 0 0 0 0 0 0 0 Coho 19 7 0 0 0 0 11 0 0 0 0 1 0 Sockeye 4 0 0 0 0 0 0 2 2 0 0 0 0

Table 5. Salmonid species sampled from resident killer whale feeding events during May-September, 2003-08. PSC Index is the total (all regions) Chinook abundance index from the Pacific Salmon Commission. Samples for October-November are not included due to the preponderance of chum salmon in the whales’ diet at that time of year, and the unequal sampling effort during these months in different years.

Year PSC Species Total Index Chinook Chum Coho Pink Sockeye Steelhead 2003 1.3 59 13 1 0 0 0 73 2004 1.16 137 17 3 0 0 0 157 2005 0.94 38 22 0 0 0 0 60 2006 0.76 66 15 0 0 2 0 83 2007 0.62 102 8 1 0 0 0 111 2008 0.68 69 13 10 1 1 1 95 Total 471 88 15 1 3 1 579

20 391

Table 6. Estimated body masses and minimum and maximum Daily Prey Energy Requirements (DPER) for individual resident killer whales, based on age- and sex-class membership.

Age- and Sex- Age Body Mass Min DPER Max DPER Class (days) (kg) (kcal/day) (kcal/day) age 1 365 465 41396 49681 age 2 730 695 55949 67146 age 3 1095 949 70650 84790 age 4 1460 1208 84645 101587 age 5 1825 1455 97359 116845 age 6 2190 1682 108510 130228 age 7 2555 1881 118014 141634 age 8 2920 2051 125941 151147 age 9 3285 2194 132447 158956 age 10 3650 2311 137720 165284 age 11 4015 2406 141944 170354 age 12 4380 2482 145303 174385 age 13, male 4745 2684 154076 184914 age 13, female 4745 2547 148143 177793 age 14, male 5110 2886 162688 195249 age 14, female 5110 2612 150970 181186 age 15, male 5475 3088 171151 205406 age 15, female 5475 2677 153779 184557 age 16, male 5840 3290 179477 215398 age 16, female 5840 2742 156571 187908 age 17, male 6205 3491 187676 225238 age 17, female 6205 2807 159346 191239 age 18, male 6570 3693 195757 234937 age 18, female 6570 2872 162106 194551 age 19, male 6935 3895 203728 244504 age 19, female 6935 2937 164850 197844 age ≥20, male ≥7300 4097 211597 253947 age ≥20, female ≥7300 3002 167562 201098

21 392

Table 7. Population DPERs (kcal/day) calculated from demographic information (2008 census) for northern resident killer whales (n = 241, aged ≥1 yr).

Min DPER Max DPER Population Population Age- and Sex-Class # whales/ class (kcal/day) (kcal/day) Min DPER Max DPER age 1 11 41396 49681 455358 546495 age 2 15 55949 67146 839230 1007197 age 3 14 70650 84790 989095 1187058 age 4 13 84645 101587 1100388 1320626 age 5 8 97359 116845 778869 934756 age 6 9 108510 130228 976594 1172055 age 7 4 118014 141634 472057 566537 age 8 4 125941 151147 503763 604589 age 9 9 132447 158956 1192026 1430604 age 10 5 137720 165284 688599 826418 age 11 8 141944 170354 1135556 1362832 age 12 5 145303 174385 726515 871924 age 13, male 3 154076 184914 462228 554741 age 13, female 1 148143 177793 148143 177793 age 14, male 2 162688 195249 325375 390497 age 14, female 5 150970 181186 754849 905929 age 15, male 0 171151 205406 0 0 age 15, female 2 153779 184557 307558 369114 age 16, male 3 179477 215398 538430 646194 age 16, female 1 156571 187908 156571 187908 age 17, male 2 187676 225238 375351 450476 age 17, female 6 159346 191239 956078 1147433 age 18, male 2 195757 234937 391514 469873 age 18, female 6 162106 194551 972635 1167303 age 19, male 2 203728 244504 407457 489008 age 19, female 2 164850 197844 329699 395687 age ≥20, male 33 211597 253947 6982705 8380261 age ≥20, female 66 167562 201098 11059076 13272498 Total Population 241 n/a n/a 34025721 40835806

22 393

Table 8. Population DPERs (kcal/day) calculated from demographic information (2008 census) for southern resident killer whales (n = 85, aged ≥1 yr).

# whales/ Min DPER Max DPER Population Population Age- and Sex-Class class (kcal/day) (kcal/day) Min DPER Max DPER age 1 3 41396 49681 124188 149044 age 2 1 55949 67146 55949 67146 age 3 4 70650 84790 282599 339159 age 4 3 84645 101587 253936 304760 age 5 4 97359 116845 389435 467378 age 6 2 108510 130228 217021 260457 age 7 2 118014 141634 236029 283269 age 8 1 125941 151147 125941 151147 age 9 0 132447 158956 0 0 age 10 2 137720 165284 275439 330567 age 11 0 141944 170354 0 0 age 12 3 145303 174385 435909 523154 age 13, male 2 154076 184914 308152 369827 age 13, female 3 148143 177793 444430 533380 age 14, male 0 162688 195249 0 0 age 14, female 1 150970 181186 150970 181186 age 15, male 3 171151 205406 513453 616218 age 15, female 2 153779 184557 307558 369114 age 16, male 3 179477 215398 538430 646194 age 16, female 0 156571 187908 0 0 age 17, male 2 187676 225238 375351 450476 age 17, female 1 159346 191239 159346 191239 age 18, male 1 195757 234937 195757 234937 age 18, female 2 162106 194551 324212 389101 age 19, male 2 203728 244504 407457 489008 age 19, female 0 164850 197844 0 0 age ≥20, male 6 211597 253947 1269583 1523684 age ≥20, female 32 167562 201098 5361976 6435150 Total Population 85 n/a n/a 12753120 15305596

23 394

Table 9. Age profile of Chinook salmon killed by northern (n = 318; predation samples collected from 1975-2008) and southern (n = 159; collected from 1974-2008) resident killer whales.

Northern Residents Southern Residents Chinook Age (years) n % n % 2 1 0.3 6 3.8 3 36 11.3 15 9.4 4 153 48.1 69 43.4 5 111 34.9 60 37.7 6 16 5.0 7 4.4 7 1 0.3 2 1.3 Total 318 100 159 100

Table 10. Average fork lengths (mm) and energy content (kcal/fish) for Chinook salmon by age-class membership (from Ford and Ellis 2006 and O’Neill et al. in prep.).

Age (year) Length (mm) Energy content (kcal/fish) 2 425 ± 1.19 1601.5 3 581 ± 2.14 4249.9 4 808 ± 3.43 11898.3 5 939 ± 4.21 19018.5 6 961 ± 15.0 20444.2

Table 11. Proportion of DPER (kcal/day) for northern resident killer whales obtained from each age class of Chinook salmon, based on an assumed diet composition of 100% Chinook. Minimum (min) and maximum (max) number of Chinook consumed was calculated based on values of kcal/fish obtained from a regression of fork length and energy content (O’Neill et al. in prep.; Table 10).

Chinook Age Min fish Max fish % of kills Min DPER Max DPER (years) per day per day 2 0.31 106,999 128,414 66.8 80.2 3 11.32 3,851,968 4,622,921 906.4 1087.8 4 48.11 16,370,866 19,647,416 1375.9 1651.3 5 34.91 11,876,903 14,254,008 624.5 749.5 6 5.03 1,711,986 2,054,632 83.7 100.5 7* 0.31 106,999 128,414 5.2 6.3 Total 100 34,025,721 40,835,806 3063 3676 *As no length data were available for 7 yr old Chinook, the 6 yr old fork length measurement was used.

24 395

Table 12. Proportion of DPER (kcal/day) for southern resident killer whales obtained from each age class of Chinook salmon, based on an assumed diet composition of 100% Chinook. Minimum (min) and maximum (max) number of Chinook consumed was calculated based on values of kcal/fish obtained from a regression of fork length and energy content (O’Neill et al. in prep.; Table 10).

Chinook Age Min fish Max fish % of kills Min DPER Max DPER (years) per day per day 2 3.8 484,619 581,613 302.6 363.2 3 9.4 1,198,793 1,438,726 282.1 338.5 4 43.4 5,534,854 6,642,628 465.2 558.3 5 37.7 4,807,926 5,770,210 252.8 303.4 6 4.4 561,137 673,446 27.4 32.9 7* 1.3 165,791 198,973 8.1 9.7 Total 100 12,753,120 15,305,596 1338 1606 *As no length data were available for 7 yr old Chinook, the 6 yr old fork length measurement was used.

Table 13. Minimum and maximum DPERs (kcal/day) supplied by Chinook salmon depending on diet composition, and the resulting numbers of Chinook per day and per year required by the northern resident killer whale population (n = 241 whales).

% Chinook in diet Min DPER Max DPER Min fish/d Max fish/d Min fish/yr Max fish/yr 100 34,025,721 40,835,806 3063 3676 1,117,832 1,341,561 90 30,623,149 36,752,226 2756 3308 1,006,049 1,207,404 70 23,818,005 28,585,064 2144 2573 782,482 939,092 50 17,012,861 20,417,903 1531 1838 558,916 670,780

Table 14. Minimum and Maximum DPERs (kcal/day) supplied by Chinook salmon (depending on diet composition), and the resulting numbers of Chinook per day and per year required by the southern resident killer whale population (n = 85 whales).

% Chinook in diet Min DPER Max DPER Min fish/d Max fish/d Min fish/yr Max fish/yr 100 12,753,120 15,305,596 1338 1606 488,453 586,215 90 11,477,808 13,775,036 1204 1445 439,608 527,593 70 8,927,184 10,713,917 937 1124 341,917 410,350 50 6,376,560 7,652,798 669 803 244,227 293,107

25 396

FIGURES

Figure 1. Ranges of northern (left panel) and southern (right panel) populations of resident killer whales. The two populations are not known to associate despite overlapping ranges.

26 397

Figure 2. Population sizes of northern (a, top panel) and southern (b, bottom panel) resident killer whales, 1974-2008.

27 398

Figure 3. Number of feeding events observed during field studies of resident killer whales, 1973-2009.

Figure 4. Monthly distribution of feeding events by northern (open bars, n = 715) and southern (closed bars, n = 222) resident killer whales.

28 399

Figure 5. Locations of 937 feeding events by resident killer whales documented during 1974-2009. Red dots indicate feeding events by northern residents, blue dots by southern residents.

29 400

Figure 6. Coastal regions in British Columbia where killer whale predation samples were collected. Numbers in parentheses indicate Pacific Fisheries Management Areas (PFMA) encompassed within each region. Sample sizes for each region are provided in Table 2.

30 401

Figure 7. Monthly distribution of salmonid species in resident killer whale feeding events, based on data provided in Table 3 (n = 806 feeding events). Sockeye and steelhead salmon are not illustrated due to their rarity in prey samples.

31 402

Figure 8. Frequency distribution of salmonid species consumed by resident killer whale in different coastal regions. Regions correspond to those shown in Figure 5, and predation data are provided in Table 2 (n = 806 feeding events). Sockeye and steelhead salmon are not illustrated due to their rarity in prey samples.

32 403

Figure 9. Locations of sampling (top), stock regions (middle), and monthly distribution (bottom) of Chinook salmon sampled from feeding events by northern resident killer whales in the northern Queen Charlotte Islands (PFMA 1). n = 33.

33 404

Figure 10. Locations of sampling (top), stock regions (middle), and monthly distribution (bottom) of Chinook salmon sampled from feeding events by northern resident killer whales in the northern mainland coast region (PFMAs 3-6). n = 52.

34 405

Figure 11. Locations of sampling (top), stock regions (middle), and monthly distribution (bottom) of Chinook salmon sampled from feeding events by northern resident killer whales in the central mainland coast region (PFMAs 7-11). n = 24.

35 406

Figure 12. Locations of sampling (top), stock regions (middle), and monthly distribution (bottom) of Chinook salmon sampled from feeding events by northern resident killer whales in the northeastern Vancouver Island region (PFMAs 12-13). n = 205.

36 407

Figure 13. Locations of sampling (top), stock regions (middle), and monthly distribution (bottom) of Chinook salmon sampled from feeding events by resident killer whales in the southeastern Vancouver Island region (PFMAs 14-19 and 28-29). Red dots indicate feeding events by northern residents, blue dots by southern residents. n = 24.

37 408

Figure 14. Locations of sampling (top), stock regions (middle), and monthly distribution (bottom) of Chinook salmon sampled from feeding events by resident killer whales in the western Vancouver Island region (PFMAs 20-24). Red dots indicate feeding events by northern residents, blue dots by southern residents. n = 136.

38 409

Figure 15. Annual indices of mortality of (a) northern and (b) southern resident killer whales and (c) abundance of Chinook salmon, 1979-2008. Deviations from an annual index value of 1 (a,b) indicate higher or lower than expected mortality rates. Annual abundance indices for Chinook salmon are from the Pacific Salmon Commission Chinook technical committee (PSC 2008).

39 410

Figure 16. Relationship between annual indices of Chinook salmon abundance and resident killer whale mortalities, 1979-2008. Killer whale mortality index values are the ratio of observed to expected deaths in the population for each year. Mortality indices are lagged one year following Chinook salmon abundance (y = -2.0412x + 3.2334, r² = 0.48673).

40 411

Figure 17. Upper and lower bound Daily Prey Energy Requirements (DPER) for male and female resident killer whales by age-class (years). Note that DPER values for whales aged 12 and under are equivalent for males and females (not differentiated by sex).

41 412

Figure 18. Lower and upper bound population daily prey energy requirements (DPERs) for northern resident killer whales (n = 241) by age- and sex-class.

Figure 19. Lower and upper bound population daily prey energy requirements (DPERs) for southern resident killer whales (n = 85) by age- and sex-class.

42 413

Figure 20. Regression of fork-length of Chinook salmon to energy density in kilocalories (from O’Neill et al., in prep.).

Figure 21. Projected increase in requirements of Chinook salmon by resident killer whales assuming a 2.6% annual population growth rate between 2008 and 2018 and diet compositions of 90%, 70% and 50% Chinook salmon.

43 414