UBC Faculty Research and Publications

Trophic Mass-Balance Model of Alaska’s Prince William Sound Ecosystem, for the Post-Spill Period 1994-1996… Okey, Thomas A.; Pauly, D. (Daniel) 1999

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-Okey_T_et_al_Trophic_mass_balance_model.pdf [ 1.22MB ]
JSON: 52383-1.0354483.json
JSON-LD: 52383-1.0354483-ld.json
RDF/XML (Pretty): 52383-1.0354483-rdf.xml
RDF/JSON: 52383-1.0354483-rdf.json
Turtle: 52383-1.0354483-turtle.txt
N-Triples: 52383-1.0354483-rdf-ntriples.txt
Original Record: 52383-1.0354483-source.json
Full Text

Full Text

i           ISSN 1198-6727  Fisheries Centre Research Reports  1999   Volume 7   Number 4        Trophic Mass-Balance Model of Alaska’s Prince William Sound Ecosystem, for the Post-Spill Period 1994-1996  2nd Edition         Fisheries Centre, University of British Columbia, Canada ii                                              ISSN 1198-6727     A Trophic Mass-Balance Model of          Alaska’s Prince William Sound Ecosystem,   for the Post-Spill Period 1994-1996         2nd Edition   Fisheries Centre Research Reports    1999 Volume 7 Number 4 iii       Trophic Mass-Balance Model of Alaska’s Prince William Sound Ecosystem, for the Post-Spill Period 1994-1996  2nd Edition         compiled and edited by  Thomas A. Okey and Daniel Pauly            published by  The Fisheries Centre, University of British Columbia  2204 Main Mall Vancouver, B.C., Canada 1998   ISSN 1198-6727 iv        AbstractInformation about the ecological components of Alaska's Prince William Sound (PWS) has in-creased considerably since the 1989 Exxon Valdez oil spill (EVOS), but the structure and func-tional characteristics of the overall food web are still not well understood. A better understanding of the whole PWS food web and its dynamics was achieved by constructing a balanced trophic model using the Ecopath approach. This was the best available framework to summarize available ecosystem information in a trophic context, as it explicitly accounts for multi-species interactions. The PWS model is a cohesive synthesis of the overall biotic community with a focus on energy flow structure, and response to perturbations--both natural and anthropogenic. Flows of biomass among the various components of the food web were quantified using estimates provided by a collaborative group of over 35 experts on PWS ecosystem components.  Forty-eight biotic components were included in the PWS model ranging from life stages of indi-vidual species to aggregated functional groups. These groups were organized into primary pro-ducers, zooplankton, benthic invertebrates, planktiverous 'forage fishes', larger fishes, birds, mammals, and detritus, for the purpose of model documentation. Estimates of biomass flows re-lated to fisheries landings and discards in Prince William Sound are also incorporated.    Biomass, production rates, consumption rates, and diet compositions were specified as (empiri-cally-based) inputs for each defined biotic component, as were migration rates, biomass accumu-lation rates, and fishery catches and discards. Outputs of the Ecopath model included biomass and flux estimates for individual groups that were refined through the collaborative mass-balancing approach, and useful characterizations of the whole food web. The outputs of Ecosim and Eco-space are also featured. These include simulations of population trajectories through time, and habitat-based re-distributions of organisms in space.  The dynamic modelling routines Ecosim and Ecospace can be used to simulate the ecosystem-level effects of disturbances and management actions, and to provide insights into ecosystem-level changes and dynamics that may occur in Prince William Sound. The Ecopath model of PWS can be used to help guide future research programs in the region, to help assess impacts of the EVOS, and to help resource agencies and local communities achieve ecosystem-based conserva-tion and management in the face of increasing human activities in the region. This approach can also be used to help distinguish the relative importance of physical forces and tropic forces in ma-rine ecosystems. An annotated list of Alutiik words was included in this volume to facilitate cross-cultural flows of ecosystem knowledge. This list might serve as one step in helping to promote a more community-based approach to management of the wild living resources of Prince William Sound. v     Director’s Foreword For many years single species stock assess-ment of fisheries has reigned supreme and separate from mainstream marine ecology, but, for marine conservation, this approach and lack of integration has been conspicu-ously unable to answer the crucial questions of our time. Such questions include the how human fisheries impact the interplay of predators, competitors and prey in natural systems, the impact, both acute and chronic, of marine pollution, and the effects of pro-gressive shoreline development on the sta-bility and value to human society of coastal ecosystems.  The first mass-balance models of marine ecosystems in the North-eastern Pacific, covering the Alaska Gyre, the shelf of southern British Columbia, and the Strait of Georgia, were constructed in November of 1996 at a workshop held at the UBC Fisher-ies Centre (see Fisheries Centre Research Report 1996, Vol. 4, No 1). That work was extended to a preliminary ecosystem model of Prince William Sound, Alaska, prior to the 1989 Exxon Valdez oil spill (see Fisher-ies Centre Report 5(2), Dalsgaard and Pauly 1997), in its most likely form between 1980-1985, based on data from published litera-ture. Ecopath models are forgiving in that they can be improved and enhanced using new information without having to be com-pletely reconstructed.  Ecopath is a straightforward trophic model-ling approach to ecosystems, which balances the budget of biomass production and loss for each component in the system by solving a set of simultaneous linear equations. The Ecopath approach is the only ecosystem model to obey the laws of thermodynamics. It is based on pioneering work by Dr J. J. Polovina from Hawai’i in the early 1980s, and was developed by Dr Daniel Pauly when he was at ICLARM, Manila, and by Dr Villy Christensen from Denmark and now at the Fisheries Centre. Dr Carl Walters at the Fisheries Centre recently developed Ecosim and Ecospace, dynamic versions of Ecopath. A Trophic Mass-Balance Model of Alaska’s Prince William Sound Ecosystem, for the Post-Spill Period 1994-1996 was published in 1998 as Fisheries Centre Research Report Vol 6 No4. This report, describing the post-spill ecosystem, builds on this earlier work by harnessing the immense body of data and information gathered during the EVOS re-search program. The 2nd Edition of this re-port improves on the previous work after feedback from the EVOS team. The model structure and parameter values have been refined after workshops and consultations. These include   explicit new components for salmon carcasses, orcas, detritus, fishery sectors and discards; improved assimilation coefficients and parameters for sharks, her-ring, orcas and others. This work represent one of the most complex mass-balance mod-els constructed to date, and moreover is sup-ported by the largest  synthesis of validated ecosystem data and research effort ever as-sembled. Simulations using ECOSIM simula-tions presented here, together with their un-certainties, are intended to receive serious consideration in the evaluation of policy op-tions for Prince William Sound. The report is the latest in a series of research reports published by the UBC Fisheries Cen-tre. A list is shown on our web site at http:/fisheries.com. The series aims to focus on broad multidisciplinary problems in fish-eries management, to provide a synoptic overview of the foundations and themes of current research, to report on work-in-progress, and to identify the next steps and ways that research may be improved. Edited reports of the workshops and research in progress are published in Fisheries Centre Research Reports and are distributed to all project or workshop participants. Further copies are available on request for a modest cost-recovery charge. Please contact the Fisheries Centre by mail, fax or email to ‘office@fisheries.com’. Tony J. Pitcher Professor of Fisheries Director, UBC Fisheries Centre vi        Table of Contents ABSTRACT.................................................................................................................................................IV DIRECTOR’S FOREWORD...................................................................................................................... V LIST OF EXHIBITS................................................................................................................................VIII PREFACE TO THE 2ND EDITION............................................................................................................ X INTRODUCTION (Thomas A. Okey).......................................................................................................... 1 Prince William Sound ............................................................................................................................ 2 The Physical Setting ........................................................................................................................................... 2 Biological Inhabitants ......................................................................................................................................... 4 The 1989 Exxon Valdez Oil Spill............................................................................................................ 5 Other Anthropogenic Stressors .............................................................................................................. 5 Natural Disturbances and Cycles........................................................................................................... 6 Defining the PWS Ecosystem ................................................................................................................. 7 Aspects of the Ecopath approach relevant to the PWS model (V. Christensen, D. Pauly, T.A. Okey).. 9 Surface areas of PWS depth zones and habitats (Tom Dean) .............................................................. 11 MODEL INPUTS........................................................................................................................................ 12 PRIMARY PRODUCERS........................................................................................................................ 12 Benthic algae and eelgrass (Tom Dean) .............................................................................................. 12 Phytoplankton (Ted Cooney and Thomas A. Okey) ............................................................................ 14 Nearshore Phytoplankton (Thomas A. Okey)...................................................................................... 15 ZOOPLANKTON .................................................................................................................................... 15 Offshore Zooplankton (Ted Cooney) ................................................................................................... 15 Nearshore Zooplankton (Robert J. Foy)............................................................................................... 18 Carnivorous Jellies  (Thomas A. Okey, Robert J. Foy, and Jennifer Purcell) ..................................... 19 BENTHIC INVERTEBRATES................................................................................................................ 19 Shallow Large Infauna (Tom Dean) .................................................................................................... 19 Shallow Small Epifauna (Tom Dean)................................................................................................... 20 Shallow Large Epifauna  (Tom Dean) ................................................................................................. 21 Small Infauna (Stephen Jewett)............................................................................................................ 22 Deep benthic groups and meiofauna  (Thomas A. Okey) .................................................................... 23 Deep large infauna ............................................................................................................................................ 23 Deep Epifauna .................................................................................................................................................. 24 Meiofauna......................................................................................................................................................... 24 PLANKTIVOROUS ‘FORAGE FISHES’ ............................................................................................... 25 Salmon Fry (Thomas C. Kline, Jr.) ...................................................................................................... 25 Adult Pacific Herring (Thomas A. Okey and Johanne Dalsgaard) ...................................................... 27 Juvenile Pacific Herring (Thomas A. Okey and Robert J. Foy) .......................................................... 28 Sandlance (Evelyn D. Brown and Thomas A. Okey)........................................................................... 29 Capelin (Evelyn D. Brown and Thomas A. Okey)............................................................................... 30 Eulachon (Evelyn D. Brown and Thomas A. Okey) ............................................................................ 30 Squid (Jay Kirsch and Thomas A. Okey) ............................................................................................. 30 LARGER FISHES.................................................................................................................................... 31 Walleye Pollock (Mark Willette).......................................................................................................... 31 Nearshore Demersal Fishes (Tom Dean) ............................................................................................ 33 Adult Salmon (Thomas A. Okey) ......................................................................................................... 34 Rockfishes (Thomas A. Okey).............................................................................................................. 36 Nearshore Rockfish (Tom Dean).......................................................................................................... 37 Miscellaneous Demersal Fishes (Thomas A. Okey) ............................................................................ 38 vii     Deep Demersals (skates and flatfishes).............................................................................................................38 Pacific Cod........................................................................................................................................................38 Lingcod.............................................................................................................................................................39 Sablefish ...........................................................................................................................................................39 Arrowtooth Flounder (Mark Willette) ................................................................................................. 40 Pacific Halibut (staff of the International Pacific Halibut Commission) ............................................. 41 Sharks (Lee Hulbert) ............................................................................................................................ 42 BIRDS ...................................................................................................................................................... 46 Invertebrate-Eating Sea Ducks (Dan Esler) ......................................................................................... 46 Seabirds and Seabird Predators (William D. Ostrand and David B. Irons) ........................................ 48 Consumption of Herring Eggs by Birds (Mary Ann Bishop and Thomas A. Okey) ............................ 52 MAMMALS............................................................................................................................................. 55 Baleen Whales (Craig Matkin and Rod Hobbs) ................................................................................... 55 Sea otter (James L. Bodkin, Dan H. Munson, and George E. Esslinger) ............................................. 55 Pinnipeds (Kathy Frost) ....................................................................................................................... 57 Orcas (Craig Matkin and Rod Hobbs) ................................................................................................. 59 Small Cetaceans (Craig Matkin and Rod Hobbs) ................................................................................ 60 DETRITUS (THOMAS A. OKEY).................................................................................................................. 61 Benthic detritus pool ............................................................................................................................ 61 Pelagic detritus pool ............................................................................................................................ 62 Input of terrestrial organic material .................................................................................................... 62 Adjusting assimilation efficiencies ....................................................................................................... 63 PWS FISHERIES......................................................................................................................................... 64 Fishery Landings Estimates for PWS, 1994-1996................................................................................ 64 Commercial landings in PWS (Thomas A. Okey) ............................................................................................64 Recreational landings in PWS (Scott Meyer)....................................................................................................64 Fishery Discard Estimates for PWS, 1994-1996 (Thomas A. Okey and Scott Meyer)........................ 65 CONSTRUCTING AND BALANCING THE PWS MODEL (Thomas A. Okey) ................................. 67 Verification of web structure................................................................................................................ 72 ECOSIM AND ECOSPACE METHODOLOGY (Thomas A. Okey) ..................................................... 73 RESULTS (Thomas A. Okey) ..................................................................................................................... 74 Temporal simulations of perturbations ................................................................................................ 78 Spatially explicit simulations ............................................................................................................... 82 ECOPATH AND RESOURCE MANAGEMENT (Thomas A. Okey) .................................................... 84 ECOSYSTEM MODELS AS CARICATURES (Jennifer L. Ruesink) .................................................... 86 AN ANNOTATED LIST OF ALUTIIQ WORDS RELEVANT TO MODELING THE PRINCE WILLIAM SOUND ECOSYSTEM (Dave Preikshot and Jeff Leer)......................................................... 89 Animals, general .................................................................................................................................. 93 Birds, general....................................................................................................................................... 93 Bird names ........................................................................................................................................... 93 Fish, general ........................................................................................................................................ 96 Fish names ........................................................................................................................................... 97 Mammal names .................................................................................................................................... 99 Invertebrates ........................................................................................................................................ 99 Plants / Protists .................................................................................................................................. 100 DISCUSSION............................................................................................................................................. 101 ACKNOWLEDGEMENTS...................................................................................................................... 103 LITERATURE CITED ............................................................................................................................ 104 viii        APPENDICES........................................................................................................................................... 116 Appendix 1. List of contributors to PWS Ecopath model ................................................................... 116 Appendix 2. Workshop Agendas......................................................................................................... 121 Appendix 3. Monthly estimates for PWS zooplankton parameters..................................................... 126 Appendix 4. Derivation of diet compositions of forage fishes ............................................................ 129 Appendix 5. Input diet compositions  of PWS animals ....................................................................... 133 Appendix 6. Diagram of arrowtooth founder spatial distribution...................................................... 135  List of Exhibits  BOX 1. BASIC EQUATIONS, ASSUMPTIONS AND PARAMETERS OF THE ECOPATH APPROACH........................... 10 BOX 2. HYPOTHETICAL ‘WHAT IF’ SCENARIOS FOR SIMULATIONS RUNS ........................................................ 73  FIGURE 1. MAP OF PRINCE WILLIAM SOUND (PWS), ALASKA. ...................................................................... 3 FIGURE 2. DEPTH CONTOURS AND DEFINED BOUNDARIES OF PWS ECOSYSTEM ............................................. 8 FIGURE 3. SEASONAL CHANGES IN PWS ZOOPLANKTON............................................................................... 16 FIGURE 4. CATCH HISTORY OF HERRING IN THE PWS AREA.......................................................................... 28 FIGURE 5. AVERAGE ANNUAL SHARK BYCATCH PER 100 HOOKS.. ................................................................ 42 FIGURE 6. BIOTIC COMPONENTS OF THE BALANCED TROPHIC MODEL OF PRINCE WILLIAM SOUND.............. 76 FIGURE 7. MIXED TROPHIC IMPACTS OF GROUPS IN THE PWS ECOSYSTEM MODEL....................................... 77 FIGURE 8. SIMULATED REMOVAL OF SHARKS FROM PRINCE WILLIAM SOUND ............................................. 80 FIGURE 9. SIMULATION OF CHANGES IN THE BIOMASS OF PRINCE WILLIAM SOUND BIOTA. ......................... 80 FIGURE 11. DIAGRAMMATIC MAP OF PRINCE WILLIAM SOUND. ................................................................... 82 FIGURE 12. SIMULATED SPATIAL RE-DISTRIBUTIONS OF THREE OF THE 48 GROUPS ...................................... 83 FIGURE 13. TEMPORAL BIOMASS TRAJECTORIES ASSOCIATED WITH SPATIAL RE-DISTRIBUTION OF BIOTA.... 84 FIGURE 14. GEOGRAPHIC RANGE OF ALUTIIQ ALASKAN YUPIK, ITS DIALECTS AND SUBDIALECTS............... 91   TABLE 1. SURFACE AREAS OF DEPTH STRATA IN PRINCE WILLIAM SOUND, ALASKA.................................... 12 TABLE 2. ESTIMATES OF THE PERCENTAGE OF SUBTIDAL HABITATS WITHIN EACH DEPTH ............................ 12 TABLE 3. RELATIVE IMPORTANCE OF HABITAT TYPE ESTIMATED FROM % OF TOTAL SHORELINE ................. 12 TABLE 4. ESTIMATES OF BIOMASS OF INTERTIDAL ALGAE IN DIFFERENT DEPTH STRATA AND HABITAT ........ 13 TABLE 5. WEIGHTED MEAN INTERTIDAL ALGAL BIOMASS............................................................................. 13 TABLE 6. SUBTIDAL MACROALGAL AND EELGRASS BIOMASS........................................................................ 13 TABLE 7. ESTIMATES OF AVERAGE BIOMASS FOR SUBTIDAL ALGAE AND EELGRASS ..................................... 13 TABLE 8. MONTHLY CHANGES IN PRODUCTION IN PWS................................................................................ 15 TABLE 9. ANNUALIZED ECOPATH PARAMETERS FOR ZOOPLANKTON. ........................................................... 17 TABLE 10. ANNUAL SUMMARY DATA FOR NEARSHORE ZOOPLANKTON ........................................................ 19 TABLE 11. BIOMASS OF SMALL INTERTIDAL EPIFAUNA IN DIFFERENT STRATA AND HABITATS...................... 20 TABLE 12. WEIGHTED MEAN BIOMASS OF SMALL EPIBENTHIC INVERTEBRATES IN THE INTERTIDAL............. 20 TABLE 13. DENSITY AND BIOMASS OF LARGE EPIBENTHIC INVERTEBRATES IN 0 TO 20 M DEPTHS ................ 21 TABLE 14. PROPORTION OF SUBSTRATE TYPES IN DIFFERENT PWS DEPTH ZONES......................................... 22 TABLE 15. ESTIMATES OF MACROBENTHIC BIOMASS..................................................................................... 22 TABLE 16. ESTIMATES OF P/B, Q/B AND DIET COMPOSITION OF MACROBENTHOS. ....................................... 23 TABLE 17 DOMINANT MACROBENTHIC GROUPS IN PWS............................................................................... 23 TABLE 18. ESTIMATED BIOMASS OF BENTHIC EPIFAUNA ............................................................................... 25 TABLE 19. DIET COMPOSITION OF PINK AND CHUM SALMON FRY.................................................................. 26 TABLE 20. ECOPATH PARAMETERS FOR SALMON FRY > 6CM......................................................................... 26 TABLE 21. ECOPATH PARAMETERS FOR SALMON FRY < 6CM......................................................................... 27 TABLE 22. ESTIMATED PRE-FISHERY HERRING RUN BIOMASS ..................................................................... 27 TABLE 23. DIET COMPOSITION OF HERRING .................................................................................................. 28 TABLE 24. FORAGE FISH BIOMASS ESTIMATES .............................................................................................. 28 TABLE 25. DIET COMPOSITION OF JUVENILE HERRING................................................................................... 29 TABLE 26. DIET COMPOSITION OF CAPELIN ................................................................................................... 29 ix     TABLE 27. DIET COMPOSITION OF SANDLANCE ............................................................................................. 30 TABLE 28. SQUID CAUGHT FROM PWS POLLOCK SURVEYS .......................................................................... 31 TABLE 29. DIET COMPOSITION OF EULACHON IN PWS.................................................................................. 31 TABLE 30. ASSIGNED DIET COMPOSITION OF PWS SQUIDS............................................................................ 31 TABLE 31. POPULATION PARAMETERS FOR WALLEYE POLLOCK.................................................................... 32 TABLE 32. SUMMER DIET COMPOSITION MATRIX FOR WALLEYE POLLOCK.................................................... 32 TABLE 33. MEAN BIOMASS PER FISH FOR NEARSHORE FISH........................................................................... 33 TABLE 34. BIOMASS OF NEARSHORE DEMERSAL FISHES BY HABITAT............................................................ 33 TABLE 35.  DENSITY OF NEARSHORE FISHES BY HABITAT ............................................................................. 33 TABLE 36. ANNUAL LANDINGS OF SALMON IN PWS. .................................................................................... 34 TABLE 37. MEAN HATCHERY AND WILD PINK AND CHUM ADULT SALMON RUNS. ......................................... 34 TABLE 38. TOTAL MORTALITY FOR FIVE SPECIES OF SALMON. ...................................................................... 35 TABLE 39. BIOMASS ESTIMATES FOR PWS ROCKFISH ................................................................................... 36 TABLE 40. DIET COMPOSITIONS OF ROCKFISH SPECIES.................................................................................. 37 TABLE 41. ESTIMATED BIOMASS OF DEMERSAL FISH SPECIES ....................................................................... 37 TABLE 42. DENSITIES OF JUVENILE COPPER ROCKFISH.................................................................................. 37 TABLE 43. REPORTED PWS COMMERCIAL LANDINGS ................................................................................... 38 TABLE 44. P/B AND Q/B VALUES FROM OTHER ECOPATH MODELS ............................................................... 38 TABLE 45. BIOMASS ESTIMATES FOR PWS PACIFIC COD............................................................................... 39 TABLE 46. DIET COMPOSITION OF PACIFIC COD............................................................................................ 39 TABLE 47.  BIOMASS ESTIMATES FOR PWS SABLEFISH. ................................................................................ 40 TABLE 48. DIET COMPOSITION OF SABLEFISHA .............................................................................................. 40 TABLE 49.  POPULATION PARAMETERS FOR ARROWTOOTH FLOUNDER ......................................................... 40 TABLE 50.  BIOMASS OF PACIFIC HALIBUT .................................................................................................... 41 TABLE 51.  SUMMER DIET COMPOSITION MATRIX FOR ARROWTOOTH FLOUNDER.......................................... 41 TABLE 52.  DIET COMPOSITION OF HALIBUT................................................................................................. 42 TABLE 53. PREY ITEMS AND % DIET COMPOSITION OF SLEEPER SHARK......................................................... 43 TABLE 54. BIOMASS AND Q/B ESTIMATES FOR PWS SHARKS....................................................................... 43 TABLE 55. DERIVATION OF ESTIMATED DIET COMPOSITION OF SALMON SHARKS.......................................... 44 TABLE 56. SPINY DOGFISH PREY COMPOSITION............................................................................................. 45 TABLE 57. GENERALIZED SHARK DIET COMPOSITION ESTIMATES ................................................................. 45 TABLE 58. ECOPATH PARAMETERS FOR INVERTEBRATE-EATING SEA DUCKS ................................................ 46 TABLE 59.  DIET COMPOSITION OF INVERTEBRATE-EATING SEA DUCKS ........................................................ 47 TABLE 60.  POPULATION ESTIMATES AND ECOPATH PARAMETER ESTIMATES FOR SEABIRDS ........................ 50 TABLE 61. DIET COMPOSITION OF SEABIRDS................................................................................................. 51 TABLE 62.  PERCENT OCCURRENCE OF PREY ITEMS OF BIRDS FROM HERRING SPAWN AREAS ........................ 53 TABLE 63. DAILY HERRING SPAWN CONSUMPTION BY AVIAN SPECIES .......................................................... 54 TABLE 64. PERCENT OCCURRENCE OF PREY ITEMS IN GUT SAMPLES OF BIRDS IN HERRING SPAWN AREAS .... 54 TABLE 65. CETACEAN INPUT PARAMETERS ................................................................................................... 55 TABLE 66.  ESTIMATES OF SEA OTTER POPULATION PARAMETERS ................................................................ 55 TABLE 67.  OTTER POPULATION ESTIMATES.................................................................................................. 56 TABLE 68. CORRECTION FACTORS FOR OTTER POPULATION ESTIMATES........................................................ 56 TABLE 69. DIET MATRIX FOR SEA OTTERS..................................................................................................... 57 TABLE 70. ECOPATH INPUT PARAMETERS FOR HARBOR SEALS IN PWS, 1992-1997...................................... 57 TABLE 71. PINNIPED DIET COMPOSITION (% WEIGHT) IN PWS, 1992-1997................................................... 58 TABLE 72.  ESTIMATED DIET COMPOSITIONS OF ORCA CATEGORIES IN PWS................................................. 59 TABLE 73. ESTIMATED MASS OF DETRITUS POOLS IN PWS............................................................................ 62 TABLE 74. PERCENT ORGANIC CARBON FROM PWS AND EARTH'S MARINE SEDIMENT .................................. 62 TABLE 75.  ESTIMATED RECREATIONAL GROUNDFISH LANDINGS.................................................................. 65 TABLE 76. ESTIMATES OF MEAN ANNUAL PWS FISHERY LANDINGS AND DISCARDS ..................................... 66 TABLE 77. PROPORTION OF SHARK DIET COMPOSITION SHIFTED TO ADULT SALMON..................................... 70 TABLE 78. BASIC INPUT PARAMETERS AND DETRITUS FATE FOR THE PWS MODEL ....................................... 71  x    ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996    Preface to the 2nd Edition Scientists and other humans must make gen-eralizations about nature because of its infi-nite complexity. It follows that any 'under-standing' of an ecosystem, or a food web, is the result of a generalization, or a rule. Such rules assume there are properties of nature that can be characterized accurately, or at least to a useful degree. Continual support of, and participation in, science and exploration is necessary because the limitations of our understanding result in imperfect characteri-zations of nature, which are then used to make decisions about human-ecosystem in-teractions. The need for understanding eco-systems increases in parallel with our ability to modify them, but our ability to understand them invariably lags behind.  In keeping with this, A balanced trophic model of Alaska's Prince William Sound eco-system for the period 1994-1996 was con-structed almost ten years after the Exxon Val-dez oil spill (EVOS) catastrophe. The main aspect of this retrospective analysis is its po-tential for enabling ecosystem-based resource planning for the future through analyses that explicitly account for multispecies interac-tions, as pointed out by Pauly et al. (1998). Its purpose, which has been achieved, was to synthesize much of the information collected since EVOS into a cohesive picture of the food web in Prince William Sound (PWS). A broad collaboration of experts from the re-gion met this goal during an iterative model construction process. In many respects, this collaborative approach gave the PWS model an aspect of self-organizing refinement that was wholly unexpected.  The initial expectations of this project were surpassed in several ways. First, virtually anyone can analyze the Prince William Sound model using the easy-to-use, win-dows-based software Ecopath with Ecosim, freely distributed on the world wide web (www.ecopath.org); Second, users can con-duct both temporal and spatial dynamic simu-lations of fishing or other disturbances; Third, natural resource management agencies, local community groups, and regional school dis-tricts are incorporating this model into their programs; and finally, models of four other aquatic ecosystems of Alaska are included along with the PWS model on a CD ROM containing useful resources relating to Alaska's aquatic ecosystems, including a da-tabase of Alaska's fishes, a dictionary of Alu-tiiq terms, videos and pictures of animals and plants, and links to additional information. The popularity of the PWS model and the positive feedback we received throughout this project almost completely drowned out the few criticisms. However, it was consideration of these criticisms that led us most directly to refinement of the PWS model. Subsequent improvements are reflected in this 2nd edition. Although the collaborators' independently-derived contributions resulted in an initial model with relatively good internal consis-tency, it was inevitable that the input parame-ters would be subject to further refinement. Refinements reflected in this 2nd edition in-clude PWS-specific estimations of the mass of nearshore and offshore detritus pools; ex-plicit treatment of fishery discards and salmon carcasses using the detritus category "nekton falls"; adjustment of assimilation efficiencies for benthic and planktonic organ-isms; explicit treatment of subsistence, rec-reational, and commercial fishery sectors; adjustment of herring catch information, ad-justment of shark consumption rates; splitting orcas into resident and transient groups, add-ing sea otters to the transient orca diet, dis-cussions of detritus pools, the dynamic nature of nearshore benthos, and the stabilizing ef-fects of complex behaviors. These refine-ments are discussed in the sections corre-ponding to the groups referred to above.  The Ecopath model of Prince William Sound is considered 'final,' only for the purposes of defining a common stopping point that can be utilized by interested parties in a format documented by this 2nd edition. However, we expect that future application of this model will result in continued refinement of its pa-rameters and structure, and our general un-derstanding of this dynamic ecosystem.   - The editors ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                 1     INTRODUCTION Thomas A. Okey Fisheries Centre, UBC, Canada The 1989 Exxon Valdez Oil Spill (EVOS) in Prince William Sound (PWS) Alaska was perceived by the media and the global public as an ecological catastrophe in light of the quantity of oil spilled (~36,000 tonnes) and the extent of its spread throughout a relatively pristine area of Alaskan coastal wilderness. In scientific terms, the scale of the disturbance was indeed catastrophic, and the impacts on biota were severe, but the full extent of the impacts remain uncertain (see Spies et al. 1996). Beyond its ecological impacts, the EVOS adversely affected native communi-ties, other local communities, fishing people, and the wider Alaskan and American public.  Determining the ecological impacts of this spill was considered necessary by resource trustee agencies and the public in order to guide cleanup and determine natural resource damages. From a scientific perspective, the EVOS was an excellent, though unfortunate, opportunity to study the ecological impacts of a large oil spill in a high-latitude marine envi-ronment. In particular, it was an opportunity to elucidate marine ecological processes re-lated to the effects of large perturbations, thereby providing insight into the structure and resilience of marine ecosystems (see Paine et al. 1996). Ideally, science programs for determining ecological impacts would quantify the ecological state before and after a perturbation at exposed and un-exposed areas (or provide experimental perturbations at realistic scales). A number of studies in-corporated such spatial comparisons, but re-search on the effects of the oil spill rarely had optimal (ideal) designs because pre-spill eco-logical information was scarce (see Hilborn 1996).  Despite these constraints, a great deal of in-formation has been collected about PWS, its ecological processes and inter-relationships, and the effects of the EVOS (Spies et al. 1996). Research programs of various scopes have collected information about particular    components and segments of the PWS eco-system revealing some mechanisms of expo-sure, effects, and ecological processes. Some segments of the ecosystem have been charac-terized at a detailed resolution, while other components and processes are more elusive, or are simply less studied. Although our knowledge of the Prince William Sound sys-tem has deepened considerably, and our knowledge of the interrelationships among ecosystem components has increased (see McRoy and Echeverria, 1990, Cooney 1997, Duffy 1997, Holland-Bartels et al. 1997), our understanding of whole-ecosystem processes can be enhanced through synthesis of existing ecosystem information. Such a synthesis was undertaken during the present project through the construction of a whole ecosystem model by a broad collaboration of experts (Appen-dix A).  The purpose of the Ecopath modelling ap-proach is to provide a cohesive picture of the PWS ecosystem by constructing a mass-balanced model of food-web interactions and trophic flows using information collected since the EVOS. This refined model was ini-tially built upon the basic PWS trophic struc-ture identified in a preliminary model of PWS (Dalsgaard and Pauly 1997). The Ecopath model includes all biotic components of the ecosystem, implicitly or explicitly, and pro-vides a quantitative description of food-web interactions and relationships, as well as en-ergy flows among components. This model not only functions as a tool for learning more about individual components, but it can help facilitate our understanding of how the sys-tem as a whole might respond to perturba-tions. To fully achieve these types of analysis and learning, Ecopath files can be used in the Ecosim and Ecospace simulations, which are temporally and spatially dynamic modelling routines that can be used to simulate indirect and whole-ecosystem effects of disturbances or management actions in both time and space (discussed later). 2    ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996    Prince William Sound The Physical Setting Prince William Sound (PWS) is a nearly en-closed embayment at the northern apex of the gulf of Alaska. At over 9,000 km2, it covers 15 times the area of San Francisco Bay and twice the area of Chesapeake Bay (Figure 1; Wheelwright 1994). PWS is a submerged section of the surrounding Chugach Moun-tains, the highest coastal range in the world, which towers up to 4 km over the waters of the sound. The depths of PWS are highly variable, to a maximum approaching 800 m, with a mean depth of 300 m  (Cooney 1993, Loughlin 1994). The coastline is highly struc-tured (Figure 1). Much of this convoluted shoreline plunges steeply to considerable depths just beyond a narrow beach shelf, or even more precipitously as vertical walls in the fjords of the western and northern sound.   PWS is located at the Northeastern corner of the Aleutian trench where the Pacific plate subducts under a bend in the North American plate making it one of the most seismically active regions in the world (Jacob 1986, Brown et al. 1989; p. 25). Ice sheets retreated from PWS 12,000 to 15,000 years ago, but the region is still shaped by its 150 glaciers. Some have begun rapid retreats, though a few advancing glaciers reflect local increases in precipitation. Some 20 of the 150 glaciers calve directly into PWS waters (Michelson 1989). The perimeter of PWS is dominated by glacially carved fjords, some with promi-nent lateral gradients of glacial sedimentation in the water column.   Much of the waters of PWS are characteristi-cally estuarine. Warm moist air arriving from the south becomes trapped, uplifted, and cooled by the surrounding Chugach Moun-tains, releasing considerable precipitation over the region. Annual precipitation ranges from 160 to 440 cm in the coastal towns of PWS, though snowfall alone can reach 2290 cm in parts of the nearby Chugach Mountains (Michelson 1989). Rain runoff and snowmelt enter from myriad streams, but icebergs and glacial melt also contribute fresh water. Even greater amounts of fresh water enter PWS as a stratified lens aloft an incurrent of marine water at the Hinchenbrook entrance. This substantial freshwater input comes from the alongshore freshwater system associated with the northwest-trending Alaska coastal cur-rent, fed by numerous rivers and glaciers from as far south as British Columbia (Wheelwright 1994). PWS thus contains complex gradients among its fresh water, es-tuarine, and marine settings.  Prince William Sound, as defined in Figure 2 by the PWS Ecopath working group (Appen-dix 1), includes a variety of deep and shallow habitats from extensive intertidal mudflats to pinnacle islands, deep basins, fjords, and holes. The annually averaged depth of the euphotic zone is approximately 25 m (D. Es-linger, UAF Institute of Marine Sciences, pers. comm.).  ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                      3      Figure 1. Map of Prince William Sound (PWS), Alaska, (modified from Braddock et al. 1996). 4    ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      The physical flux of Prince William Sound carries over into the biological regime. The profusion of its wildlife--the aggregations of seabirds, sea otters, salmon, herring, killer whales, periwinkles, jellyfish, ea-gles--tends to obscure the variability of the numbers within the species.       Jeff Wheelwright, Degrees of Disaster, 1994  Biological InhabitantsPrince William Sound, like other marine ecosystems, is characterized by a wide vari-ety of plants and animals distributed un-evenly in space and whose populations fluc-tuate in response to physical and oceano-graphic changes, occurring over a range of scales. Some organisms are adapted to un-dergo considerable fluctuations over time, like krill whose populations can vary by a factor of 50, while other organisms have developed more stable life histories through mechanisms such as prey switching, food storage, and mobility. The input of solar and imported energy into the system is mediated by independently varying physical cycles and disturbances, but the flow of that energy through the biotic components of the system is stabilized not only by the species-level mechanisms mentioned above, but also by community-level mechanisms such as op-portunistic and competitive compensations within the ecosystem’s food web. In this way, variability and shifts in populations can effectively ‘even out’ the energy flow through the system in the face of environ-mental disturbances and physical fluctua-tions.  Thus, ecosystems contain both highly vari-able and less variable components as well as a tendency for dampening of energy throughflow, through individual, population, and community level compensations. Not-withstanding such biotic ‘stabilizing’ mechanisms, or the importance of physical changes and disturbances, constraints in en-ergy flow (feeding) exist throughout the sys-tem such that organisms in the food web must eat enough of the appropriate foods to sustain themselves, and the population levels of prey are somewhat controlled by their predators. Alternatively, these feeding con-straints are lifted by the extent of feeding plasticity—the organism’s proclivity for prey switching. The static Ecopath modelling approach en-ables a description of the possible scenarios of relationships, flows, and interactions based on the known conditions in an ecosystem during a particular time period. The dynamic Ecosim approach, which then follows, enables simula-tion of particular disturbances or agents of physical forcing on the system or on particular biotic components within the context of an interactive ecosystem, based on the known interactions and energy flow constraints. Moreover, Ecosim can be re-expressed in a spatial context, leading to a spatially-explicit routine called Ecospace. The overarching question of the EVOS Re-search program is also the most persistent question of the general public: “What are the short-term and long-term effects of the Exxon Valdez Oil Spill?”  The state or trajectory of a biological community is controlled by cyclic and other changes in the physical world as well as trophic interactions and constraints. Based on this notion, the ecological effects of EVOS can be understood best if examined within the contexts of known physical stress-ors, both natural and anthropogenic, and whole inter-connected communities. But be-cause temporal and spatial controls (compari-sons) were virtually unavailable after the spill, new analytical tools are needed to describe the interrelationships, constraints, and trajectories of the PWS biotic community. An empirically based, mathematical matrix describing these interactions—the PWS Ecopath model—can be used to reveal indirect and whole ecosys-tem effects, to the extent that input estimates are accurate. Furthermore, the relative influ-ence of various physical and biotic factors contributing to the state of the ecosystem can be isolated within the analysis, to the extent that effects of the various factors are under-stood. Factors known to influence the marine ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                      5 community structure of PWS are listed be-low: The 1989 Exxon Valdez Oil Spill The supertanker T/V Exxon Valdez ran aground on Bligh Reef, in northeastern Prince William Sound on the night of March 24, 1989, after an inexperienced helmsman and other members of its small, sleep-deprived crew steered outside normal ship-ping lanes in Valdez Arm to avoid small icebergs calved by the nearby Columbia Glacier (Figure 1). This fateful maneuver was common practice in this dangerous and unmonitored passage with the goal of reach-ing the high seas more quickly. The stranded vessel spilled over 40 million liters (36,000 tonnes) of crude oil--the largest oil spill in United States history--into a largely pristine coastal marine environment renowned for its abounding wildlife, productive fisheries, and majestic scenery. During the weeks follow-ing the spill, the oil spread throughout cen-tral and southwestern PWS, into the Gulf of Alaska, along the Kenai Peninsula, into Cook Inlet, to Kodiak Island, and along the Alaska Peninsula (see Figure 2 in Paine et al. 1996).  The initial effects of the oil spill were catas-trophic in magnitude and scale, killing per-haps several hundred thousand seabirds and other birds of 90 species; 3,500-5,500 sea otters; 1,000 bald eagles, 10% of the world’s population of the threatened Kittlitz’s mur-relet; a number of killer whales; 300 harbor seals, other marine mammals; and vast numbers of fish, invertebrates, and plants (Loughlin 1994, Paine 1996, Spies et al. 1996). Although the spill’s full acute effects will remain uncertain, there is little doubt that they cascaded and reverberated through the impacted systems in addition to operat-ing directly and immediately. Even more uncertain than initial impacts are questions of recovery, resiliency, and long-term ef-fects, which the current study can help re-veal.  Other Anthropogenic Stressors Fishing - Fishing in PWS is conducted by subsistence, commercial, and recreational user groups. A variety of target species and gear types are used by these user groups, resulting in a broad regime of fisheries exploitation that tends towards efficiency through opportunistic adaptations to changes in the relative abun-dances of prey, as to predators. There is evi-dence that aboriginal communities of the Pa-cific Coast of North America had considerable impacts on coastal and marine biological re-sources such as shellfish, salmon, and sea ot-ters/kelp forest communities prior to the arri-val of Russians and other Europeans during the latter part of the Millennium. However, there is also evidence of sustained aboriginal use of Pacific marine food resources for at least 4,000 years (McEvoy 1986). Relatively consistent interactions between native com-munities and marine resources over such time periods can lead to co-evolutionary adapta-tions, and thus a natural resiliency of these marine populations to aboriginal fishing stresses.  Modern commercial, recreational, as well as subsistence fishing (currently, any Alaskan resident can fish for ‘subsistence’) are new, ‘exotic’, disturbances in character and magni-tude to which ecosystem components are not directly adapted. Recent severe declines in king crab and shrimp in PWS, and throughout Alaskan waters, coincident with intensive fishing practices (NMFS 1996), are possible examples of modern fishery impacts on re-gional populations which undergo natural cy-cles of abundance and resiliency in response to oceanographic cycles. Thus, an analysis of the impacts of EVOS should also consider ecological changes related to fishing.     Forestry - Forestry practices influence aquatic systems, as watersheds and sediment regimes are modified. Forestry-related impacts include increased sedimentation in rivers and estuar-ies, modified hydrodynamic regimes including flow rates and temperatures, disrupted nutrient cycling, modified micro- and meso-scale cli-mate patterns, and disrupted habitats of ma-rine/terrestrial cross-over species (e.g., mur-6     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996    relets, salmon, eagles, mink), and other spe-cies that indirectly influence marine ecosys-tems (e.g., bears).  Prince William Sound is almost entirely sur-rounded by the Chugach National Forest and native or other private lands on which log-ging is conducted (EVOSTC 1994, CNF 1998). Commercial timberlands comprise 31% of the 1,105,000 acres of forested land managed by the Chugach National Forest, and 8.5% of the total forested land is cur-rently considered to be a suitable land base for timber production (CNF 1998). Prince William Sound contains 65% of these 94,000 suitable acres1 in its watershed, and virtually all of this area slopes towards the waters of the sound. Methods of cutting in-clude tractor, cable, and helicopter logging. Although allowable logging rates have not been realized on CNF-managed lands, and ecological impacts to the sound are thus probably limited at present, a substantial proportion of the forested slopes adjacent to the sound’s waters have been identified as tentatively suitable for logging, in addition to the land that is already designated suitable (CNF 1998). At a 120 year rotation rate, PWS timber outputs would be worth over 2 million (1978 dollars) annually. Future changes in forestry practices are uncertain, but management latitude exists to accom-modate industry-driven increases in logging leading to adverse impacts in PWS waters and coastlines.  Coastal Development - The shorelines of PWS are undeveloped relative to other areas of the coastal U.S. However, there are three small port cities, five hatcheries, fourteen seafood processing operations, and several native communities. Each of the three cities harbors a fleet of fishing boats whose pres-ence has localized impacts on marine com-munities (see Lenihan et al. 1990), but each of the cities has its own unique impacts on PWS. McRoy (1988) and Feder and Jewett (in McRoy 1988) described direct effects of construction of oil terminal facilities on ben-thic communities in Port Valdez. Coastal                                                         1 1 acre = 4,047 m2 development and other uses may increase es-pecially now that a road from Anchorage to Whittier is slated for completion in 1999.  Pollution - Sources of pollution in PWS other than major oil spills include urban runoff, municipal outfalls, industrial outfalls (includ-ing cannery waste) and oil terminal opera-tions, hatchery-associated pollution, fishing boats, and shipping operations including bal-last water effluents. These sources contribute pollutants to the PWS ecosystem that are less obvious than large oil spills, but which are more continuous and represent chronic stresses. These patterns of coastal develop-ment and pollution, while unable to mask eco-logical signals associated with the EVOS, may be able to mask chronic stressors, thus proba-bly leading to confounding signals in some areas.  Natural Disturbances and Cycles  Earthquakes - The effects of Earthquakes and associated seismic sea waves can be severe in Prince William Sound. Though infrequent on human time scales, these disturbances may occur regularly in the region, corresponding with cycles of crustal stresses and releases. Understanding earthquake effects is important for understanding the effects of EVOS be-cause the strongest earthquake ever recorded in North America (9.2 on the Richter scale) occurred in PWS just 25 years before EVOS, on March 27, 1964. As for the EVOS, the ef-fects of the earthquake on intertidal ecosys-tems were conspicuous and severe, whereas the effects on subtidal ecosystems were incon-spicuous and uncertain. PWS tilted and elon-gated southward (by as much as 18 metres) during the 1964 earthquake, submerging some intertidal communities below the tides while stranding others nine metres above the tides (Wheelwright 1994). The ecological effects of the earthquake were exemplified by an 11 to 40% destruction of commercially important bivalves and a 90% destruction of mussels (NAS 1970 in Wheelright 1994). The quake likely caused widespread slumping and debris flows in the deeper subtidal (McRoy 1988), ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                      7 whereas the high pressure spray cleanup af-ter EVOS re-distributed beach sediment downslope to the shallow subtidal (J.L. Rue-sink, University of Washington, Zoology, pers. comm.). Long-term studies of recovery were not conducted after the earthquake (Wheelwright 1994), but quake-related eco-logical changes might have been ongoing when the EVOS occurred 25 years later.       Sea otters - Sea otters began expanding back into their historic range sometime after the signing of the International Fur Seal Treaty of 1911. Sea otters are known to have con-siderable influence on the structure of both hard and soft-bottom nearshore marine communities in Alaska (Estes et al. 1974, Estes and Duggins 1995, Kvitek and Oliver 1992, Kvitek et al. 1992), and they may have been still expanding and increasing in PWS when the EVOS occurred (also see McRoy 1988). Because of the broad eco-logical influence of this species, the studies of EVOS impacts should be interpreted in light of the changes and status of sea otter populations.    Atmospheric and Oceanographic Cycles - Atmospheric and oceanographic cycles oc-curring on various time scales can force, or influence, components of ecosystems in the Gulf of Alaska, and PWS. These include ENSO events (3-7 year period), ‘regime shifts’ in atmospheric pressure patterns and storm tracks (10+ year period), the effect of lunar declination on ocean temperatures (18.6 year period), and atmospheric changes caused by sunspots (11 years) (NRC 1996). It is possible that some ecological signals associated with these forcing mechanisms are separate from ecological changes that occurred coincident with the EVOS, but the interactions of these various forcing mecha-nisms could result in non-cyclic, or chaotic, physical and ecosystem trajectories (Parker et al. 1995). One notable event that had the potential of confounding ecological signals of an oil spill was unusually cold weather in the winter of 1989 (Wheelwright 1994). Such an event has the potential to cause un-usual stress to intertidal communities. Although changes in the PWS ecosystem are undoubtedly influenced by these and other natural cycles and mechanisms these changes are not easily predictable given our current level of knowledge about PWS and the sur-rounding GoA. Elucidating the effects of a strong event such as EVOS, much less pre-dicting the effects of such a disturbance, is challenging. Nevertheless, trophic constraints exist even in such dynamic ecosystems, and examining these constraints using tools such as Ecopath may lead to a better understanding of indirect, or ecosystem-level, responses.  Defining the PWS Ecosystem The first necessary step to constructing an Ecopath model is to define the ecosystem to be modeled. Although no ecosystem on earth is self-contained or truly separate from other ecosystems, it is useful to define distinct eco-systems. Some ecosystems are naturally well defined or distinct based on characteristics such as geography, climate, oceanography, or biotic distributions (see Defining the Ecosys-tem in NRC 1996). The Prince William Sound ecosystem is relatively easy to define, as it is somewhat separated from the Gulf of Alaska by Hinchenbrook Island, Montague Island, and other islands and peninsulas (Figure 1 and Figure 2).  Some PWS organisms spend their entire life cycles inside the Sound; others reside there for only part of their life cycles, migrating in and out of adjacent rivers, the Gulf of Alaska, or to and from distant latitudes. Still, the PWS Ecopath working group agreed on the bounda-ries of the PWS ecosystem presented in Figure 2, though some inevitable limitations of these boundaries were noted by some participants (e.g. some small pelagic fishes spawn on both sides of Montague Island, and the outside of Hinchenbrook Entrance is particularly unique and productive; E. Brown, UAF Institute of Marine Sciences, pers. comm.).      8     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                    Figure 2. Depth contours and defined boundaries of PWS ecosystem (GIS analysis and mapping provided by G. Esslinger, Alaska Biological Science Center, USGS). Prince William Sound ecosystem boundaries are delineated by the edge of the colored areas and were defined and agreed on by the PWS Ecopath work-ing group. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                      9 Aspects of the Ecopath approach and software relevant to the PWS model Villy Christensen, Daniel Pauly, and  Thomas A. Okey UBC Fisheries Centre, Vancouver BC  The Ecopath model was originally described J. Polovina (1984, 1995) of the U.S. Na-tional Marine Fisheries Service (Honolulu Laboratory). V. Christensen and D. Pauly, previously both at the International Center for Living Aquatic Resources Management (ICLARM), carried further the work (see Christensen and Pauly 1992a), and made it widely available in the form of a well-documented software for computers running MS-DOS  (Christensen and Pauly 1992b), and later Windows (Christensen and Pauly 1995, 1996). Both versions allow rapid con-struction and verification of mass-balance models of ecosystems, as is clear from its present (1999) distribution to 1600 regis-tered users in more than 90 countries.  The data requirements of an Ecopath model are expressed by its two ‘Master Equations’. These equations are based on an assumption of mass-balance, and formulate that for any given group its production can be described as:      …1 and further that …2 The first Master Equation is crucial in link-ing predator and prey in a system. Re-expressed and -arranged the equation reads,  …3 where Bi and Bj are biomasses (the latter pertaining to all consumers of i);   P/Bi is the production/biomass ratio, equiva-lent to total mortality (Z) under most circumstances (Allen 1971); EEi is the ecotrophic efficiency, or the frac-tion of production (P= B*(P/B)) that is utilized within the system (including net migration and biomass accumula-tion);  Yi is equal the fisheries catch per unit area and time (i.e., Y = F*B);  Q/Bj the food consumption per unit biomass of j; and  DCji the contribution of i to the diet of j (see also Box 1); BAi is the biomass accumulation of I (posi-tive or negative, flow rate with units of energy per unit area and time); NMi is the net migration of I (emigration less immigration) with unit of energy per unit area and time.  An important aspect facilitating construction of an Ecopath model is that P/B under most circumstances corresponds to total instanta-neous mortality rate (Z) in most circum-stances (Allen 1971). There are several ways to estimate production (and P/B) directly, however, the combination of cohort-specific abundance and growth data required for many of these methods is usually difficult to assemble. Thus, Allen's formal demonstra-tion of the relationship between P/B with Z is extremely useful, as numerous methods exist, in fisheries science for the estimation of Z from catch-at-age (Ricker 1975), length-frequency (Pauly and Gayanilo 1997), or other data (Pauly 1984).  An attribute of the Ecopath approach is that all of the parameters of its first master equa-tion are amenable to direct estimation, ex-cept the ecotrophic efficiency, which is thus often left as the unknown to be estimated when the master equation is solved.  The steps involved in construction of an Ecopath model consist essentially of: (i) Identification of the area and period for Production = Catches + Predation + Biomass accumula-tion + Net migration + Other mortality, Consumption = Production + Unassimilated food + Respiration. Bi ⋅ (P/B)i ⋅ EEi  = Yi + Σ Bj ⋅ (Q/B)j ⋅ DCji + BAi + NMi 10     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996    which a model is to be constructed; (ii) Definition of the functional groups (i.e., ‘boxes’) to be included; (iii) Entry of a diet matrix, expressing the fraction that each ‘box’ in the model represents in the diet of its consumers; (iv) Entry of food consumption rate, of production/biomass ratio or of biomass, and of fisheries catches, if any, for each box; (v) Balance the model, or modify entries (iii and iv) until input = output for each box; (vi) Analyze model outputs (e.g., system characteristics, estimated trophic levels) including simulations of functional re-sponses to disturbances and other changes;Box 1. Basic equations, assumptions and parameters of the Ecopath approach  The mass-balance modelling approach documented in this report combines an approach by Polovina and Ow (1983) and Polovina (1984, 1985) for estimation of biomass and food consumption of the various elements (species or groups of spe-cies) of an aquatic ecosystem (the original ‘Ecopath’) with an approach proposed by Ulanowicz (1986) for analysis of flows between the elements of ecosystems. The result of this synthesis was initially implemented as a DOS software called ‘Eco-path II’, documented in Christensen and Pauly (1992a, 1992b), and more recently in form of a Windows software, ‘Ecopath 3.1’ (Christensen and Pauly 1995, 1996) and Ecopath with Ecosim (Pauly 1998, Walters et al., in press). Unless noted oth-erwise the word ‘Ecopath’ refers to the latter, Windows version. The ecosystem is modeled using a set of simultaneous lin-ear equations (one for each group i in the system), i.e. Production by (i) - all predation on (i) - nonpredation losses of (i) - export of (i) = 0, for all (i). This can also be put as  Pi-M2i - Pi (1-EEi) - EXi = 0                                   …1)  where Pi is the production of (i), M2i is the total predation mortality of (i), EEi is the ecotrophic efficiency of (i) or the pro-portion of the production that is either exported or predated upon, (1-EEi) is the “other mortality”, and EXi is the export of (i). Equation (1) can be re-expressed as   Bi*P/Bi - ΣjBj*Q/Bj*DCij-P/Bi*Bi(1-EEi)-EXi =0           …1) or Bi*P/Bi*EEi - ΣjBj*Q/Bj*DCij - EXi = 0                                     ...2)  where Bi is the biomass of (i), P/Bi is the production/biomass ratio, Q/Bi is the consumption/biomass ratio and DCij is the fraction of prey (i) in the average diet of predator (j). Based on (2), for a system with n groups, n linear equations can be given in explicit terms:  B1P/B1EE1 - B1Q/B1DC11-B2Q/B2DC21 - ...-BnQ/BnDCn1 - EX1 = 0  B2P/B2EE2 - B1Q/B1DC12 - B2Q/B2DC22 - ...-BnQ/BnDCn2 - EX2 = 0  BnP/BnEEn - B1Q/B1DC1n - B2Q/B2DC2n - ...-BnQ/BnDCnn - EXn = 0 This system of simultaneous linear equations can be solved through matrix inversion. In Ecopath, this is done using the generalized inverse method described by MacKay (1981), which has features making it generally more versatile than stan-dard inverse methods. Thus, if the set of equations is over-determined (more equations than unknowns) and the equations are not consistent with each other, the generalized inverse method provides least squares estimates which minimize the discrepancies. If, on the other hand, the system is undetermined (more unknowns than equations), an answer that is consistent with the data (al-though not unique) will still be output. Generally only one of the parameters Bi, P/Bi, Q/Bi, or EEi may be unknown for any group i. In special cases, however, Q/Bi may be unknown in addition to one of the other parameters (Christensen and Pauly 1992b). Exports (e.g., fisheries catches) and diet compositions are always required for all groups. A box (or “state variable”) in an Ecopath model may be a group of (ecologically) related species, i.e., a functional group, a single species, or a single size/age group of a given species. A term for biomass accumulation (Bacc) may be added to equa-tion (1) in cases where biomass is known to have changed over the period considered in the model. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                      11 These steps can be easily implemented if ba-sic parameters can be estimated (see also Box 1), especially as numerous well-documented examples exist of Ecopath ap-plications to aquatic ecosystems (see Pauly and Christensen 1993, and contributions in Christensen and Pauly 1993, and Pauly and Christensen 1996). We sometimes refer here to three ecosystems that have much in com-mon with PWS (the Strait of Georgia, the coast of British Columbia, and the Alaska gyre), documented through the contributions in Pauly and Christensen (1996). In the pre-sent report, details are provided, by func-tional group, on how items (ii) to (vi) were implemented, for the period (1994-1996) in a defined PWS (Figure 2).  Construction of an Ecopath model is fol-lowed by model balancing. The first law of thermodynamics states that energy is neither created nor destroyed, but changed from one form to another. The total energy in a closed system remains constant, though the form of that energy changes. The PWS Ecopath model is not based on the assumption of a closed system; rather, the working assump-tion is that the energy flowing into PWS’s biotic system (primary production, and im-ported secondary production) is equal to the energy used within the defined system and flowing out of it.  For the purposes of the model, the assump-tion of mass-balance (conservation of en-ergy) is also made for every identified com-ponent of the ecosystem. However, the Eco-path formulation includes a biomass accu-mulation factor so that trends in populations, or ecosystem components, can be repre-sented, and hence the model is not necessar-ily a steady-state model. The assumption of mass-balance is extremely useful for param-eterization of ecosystem models, we always have imperfect knowledge, and mass-balance offers a powerful constraint to the parameterization process. An iterative model balancing approach can serve to increase knowledge about ecosystem components as well as the whole ecosystem, especially if conducted within a collaborative synthesis of information (see the following section on collaboration). This is so because energy flows in and out of each component must be reconciled among connected components. The balancing methodology employed for the PWS model is described in the section on ‘Constructing and Balancing the Model’ fol-lowing the ‘Model Inputs’ section below.  Contributed diet compositions and the overall food web produced by the model were com-pared to the food web elements previously published for PWS by McRoy and Wyllie Echeverria (1990). This procedure for verifi-cation was conducted for every component of the model, and comments regarding similarity are included at the end of this report.  The current project also features examples of the uses of Ecosim and Ecospace to simulate spatial and temporal responses of biotic com-ponents to various perturbations and scenar-ios. This is followed by a discussion of the application of the PWS model for future re-source planning, and how it may shed light on the effects of EVOS.     Surface areas of PWS depth zones and habitats Tom A. Dean Coastal Resources Associates Vista, California, USA Estimating biomasses on a Sound wide basis requires estimating the areal extent over which organisms are distributed. For example, sampling in the nearshore is often stratified by depth, habitat type, or both. As a result, rais-ing local estimates to the sound as a whole requires estimation of the relative proportions of each depth or habitat type within the eco-system. Estimates of area covered by different depth strata are given in Table 1. Estimates of areas covered by different subtidal habitats are given in Table 2, and areas covered by differ-ent intertidal shoreline types are given in Ta-ble 3.    12     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996       MODEL INPUTS The ecosystem components in the 1994-1996 PWS model are organized into groupings de-fined by both trophic and taxonomic consid-erations. Within these groupings, components are presented in order of descending trophic level, as estimated by the Ecopath model.  PRIMARY PRODUCERS  Benthic algae and eelgrass Tom A. Dean Coastal Resources Associates Vista, California, USA In Prince William Sound, the nearshore zone, from the upper intertidal (approxi-mately +3 m above mean lower low water) to depths of approximately 20 m, is vegetated by seaweeds and eel-grass. Fucus gardneri dominates in the intertidal zone while a variety of kelps (Agarum cribrosum, Laminaria spp., and Nereocystis luetkeana) and eel-grass (Zostera marina) are dominant in the subtidal zone. Intertidal algal biomass in western Prince William Sound was estimated by Highsmith et al. (1994) in 1990 and 1991 following the EVOS using a stratified random sampling design. Sampling was conducted at 3 depth strata within 5 habitat types at both oiled and unoiled sites. Estimates of den-sity and biomass of subtidal algae and eelgrass were made by Dean et al. (1996a, 1998) using a similar stratified random sampling design. We use values from control loca-Table 1. Surface areas of depth strata in Prince William Sound, Alaska. Depth stratum (m) Area (km2) % of area Intertidal +3 to 0  300 a 3.31 0 to 10  709 b 7.83 10 to 20  709 b 7.83 20 to 100  2,018 b 22.28 > 100 5,323 b 58.76 Totals 9,059 b 100.00 a. Estimated based on an average 20 m per 1 m of vertical in the intertidal zone, based on data of the Alaska Department of Natural Resources (1991) and unpublished measurements (T. A. Dean and S. Jewett). b. Based on GIS analysis of NOAA bathymetric data by G. Esslinger, and unpublished by T. A. Dean and S. Jewett Table 2. Estimates of the percentage of subtidal habitats within each depth in Prince William Sound, Alaska. Estimates for < 20 m are based in part on unpublished side-scan sonar records of substrate type. Depth range (m) Habitat type  2-10 11–20 20 -100 >100 Hard Substrate 85 66 30 10    Bays 42 33 - -    Points 42 33 - -    Nereocystis 1 0 - - Soft Substrate 15 33 70 90    Eelgrass 15 0 - -    No vegetation 0 33 - - Table 3. Relative importance of habitat type estimated from % of total shoreline (from Sundberg et al. 1996). These estimates do not account for possible differences in beach widths between habitats.  Shoreline type % in PWS Habitat type Habitat % Exposed rocky 13 exposed rocky - Exposed wave-cut platforms 11 exposed rocky 23 Find sand beaches 1 fine textured beaches - Coarse sand beaches 0 fine textured beaches 1 Mixed sand/gravel beaches 21 coarse textured beaches - Gravel/cobble/boulder 20 coarse textured beaches - Exposed tidal flats 0 coarse textured beaches 41 Sheltered rocky 30 sheltered rocky 30 Sheltered tidal flats 3 estuarine - Marshes 2 estuarine 5 Total 100 All 100 ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     13 tions to estimate pre-spill conditions and assume that they are representative of the entire Prince William Sound for the decade from 1980 to 1989, prior to the oil spill. The dominant alga in the intertidal was Fu-cus gardneri, comprising over 90% of the biomass. The Spring (May) biomass of algae was highest on rocky shores (sheltered rocky and exposed rocky habitats) and was gener-ally higher in the mid and lower tide zones (Table 4). The weighted mean (based on proportional coverage by each habitat type) biomass density was estimated at about 1,058 t ww km-2 (Table 5). Expressed on a Sound wide basis, this is equivalent to 35 t km-2. Agarum cribrosum and Laminaria saccharina were the dominant subtidal macroalgae in sheltered bays (Dean et al. 1996a). Generally, these two species constituted more than 90% of total macroalgal biomass. Agarum cribro-sum also dominated on exposed points (more than 60% in terms of number of individuals). Less abundant algae were Laminaria sac-charina and L. bongardiana (= groenlandica). Nereocystis habitats are located on exposed sites, and the algal diversity was higher than in the other two habitats. The kelp forest structure at these locations consists of a can-opy of Nereocystis luetkeana with an under-story of L. bongardiana (61% of the biomass), L. yezoensis, Pleurophycus gardneri, and A. cribrosum. Eelgrass dominated in shallow waters (less than 5 m) in bays, generally at stream mouths. Biomass estimates from different habitats are given in Table 6. The biomass estimate, weighted by proportion of each habitat in Table 4. Estimates of biomass (t⋅ww⋅km-2) of intertidal algae in different depth strata and habitat types of PWS (from Highsmith et al. 1994). Depth strata are as follows: MVD1 = high, intertidal, + 2.0 to + 3.0 m, MVD2 = mid intertidal, + 1.0 to + 2.0 m, and MVD3 = low intertidal, 0 to + 1 m (from Highsmith et al. 1994). Habitats are as defined in Table 4. Weighted mean intertidal algal biomass (t ww km-2) (from Highsmith et al. (1994). Habitat type Depth stratum Biomass (t⋅ww⋅km-2) May-90 Biomass (t⋅ww⋅km2) May-91 Mean high 918 1,184 mid 1,899 1,705 low 2,340 1,266 Sheltered rocky Average 1,719 1,385 1,552 high 80 136 mid 665 490 low 640 820 Coarse textured Average 462 482 472 high 364 438 mid 471 1,634 low 157 657 Estuarine Average 331 910 620 Exposed  high 822 1,026  mid 1,672 1,692  low 1,351 3,024  rocky Table 5. Weighted mean intertidal algal biomass (t ww km-2) (from Highsmith et al. (1994). Habitat type Biomass density (t⋅km2) % of area Overall biomass in PWS Sheltered rocky 1,552 30 465.6 Coarse textured 472 41 193.5 Estuarine 620 5 31.0 Exposed Rocky 1,598 23 367.5 Fine Textured 0 1 0 All 4,242 100 1,057.7 Table 6. Subtidal macroalgal and eelgrass biomass (t⋅ww⋅km-2) estimates for different Prince William Sound, Alaska habitats (from Dean et al. 1996a and 1998). Habitat Biomass density (t⋅km-2) (2-10 m) Biomass density (t⋅km-2) (11-20 m) Bays 1,766 529 Points 2,690 678 Nereocystis 6,240 0 Eelgrass 1,232 0 No Vegetation 0 0 Table 7. Estimates of average biomass (t ww km-2) for subtidal algae and eelgrass in Prince William Sound, Alaska (from Dean et al. 1996a and 1998). Habitat Biomass density (t⋅km-2) % by area Biomass (t) Bays Shallow 1800 21.0 378 Bays Deep 530 16.5 87 Points Shallow 690 21.0 145 Points Deep 680 16.5 112 Nereocystis Shallow 6200 0.5 31 Nereocystis Deep - - - Eelgrass Shallow 1200 7.5 90 Eelgrass Deep - - - No Veg Shallow - - - No Veg Deep - 33.0 - Total - - 843 14     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Prince William Sound, is 844 tonnes t⋅ww⋅km-2 (Table 7). Expressed on a Sound wide basis, this is equivalent to 132 t⋅km-2. The total biomass estimate for benthic algae and eelgrass, for both intertidal and subtidal habitats, is 167 t⋅km-2. The biomass estimates are based on observa-tions made in summer (May through Sep-tember). However, we know that in winter months, biomasses of both algae and eel-grass are reduced. Based on some very pre-liminary estimates given by Rosenthal et al. (1977), we estimate that the winter biomass is about 50% of the summer standing stock. The average of the winter and summer bio-mass estimates would be 125.25  t⋅km-2, which can be used as an annual average biomass of macroalgae and eelgrass. The P/B ratio for algae and eelgrass is about 4 year-1, based on estimates of algal and eel-grass P/B ratios given in Luning (1990). It is assumed that about 1% of algal and eel-grass production is consumed by herbivores, about 15% is exported as dissolved organics, and about 84% goes to detritus. These are similar to values given by Luning (1990) ex-cept that the percentage grazed is about 10% of that suggested by Luning. There are few grazers in Prince William Sound (especially urchins) compared with other kelp communi-ties. Most of the grazing is by small epifaunal invertebrates (especially amphipos, gastro-pods, and crabs), and large epifaunal inverte-brates (mostly crabs and sea urchins). Some kelp is harvested as part of the herring roe on kelp fishery. This is localized in her-ring spawn areas and thought to be a very small portion of the yearly production of kelp. No kelp landnings or discards were specified, since the roe-on-kelp fishery was closed dur-ing the modeled period (1994-1996).   Phytoplankton R. Ted Cooney Institute of Marine Sciences University of Alaska, Fairbanks, USA Thomas A. Okey Fisheries Centre, UBC, Canada Phytoplankton is the main source of annual primary production in PWS, and its main components are diatoms and phytoflagellates. These organisms form the base of this marine food web, as they turn solar energy into chemical energy. Diatoms are photosynthetic, single-celled protists (division Chrysophyta) whose identifying characteristic is silicified cell walls making up the lid-like valves of a protective frustule (Wetzel 1983). Phytoflag-ellates are the autotrophic (photosynthetic) group of the protozoan subphylum Mastigo-phora. They usually possess flagella for loco-motion, but unlike their heterotrophic coun-terparts, they contain chlorophyll and are treated as algae by phycologists (Barnes 1987). Abundances (densities) of these two A note on nearshore groups Tom A. Dean Coastal Resources Associates Vista, California, USA Summaries are provided for several nearshore (depths less than 20 m) groups of organisms in PWS, Alaska. These groups include benthic algae and eelgrass, shallow small epibenthos, shallow large infauna, shallow large epibenthos, and near-shore demersal fish. The rigor used in deriving these estimates varied by group, and there are still considerable data that could be mined to refine estimates. I have indicated potential sources of data that I am aware of, but have not had time to explore. No confidence inter-vals are given, but could be obtained with more work. Estimation of confidence intervals is com-plex and would likely require simulation. It is important to note that most of the data that serve as the basis for these estimates are from the western portion of Prince William Sound, and this likely produces several biases. The eastern portion of the Sound is shallower and has a higher propor-tion of soft substrates. Therefore, it is likely that we have overestimated algal and epifaunal biomasses and underestimated infaunal biomass. It also likely that fish assemblages are quite different in the East-ern and Western Sound, leading to potential biases in estimated nearshore demersal fish biomass. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     15groups vary considerably over the course of a year with diatoms blooming in the spring and flagellates numerically dominating in the winter and late summer; both groups are more-or-less equally abundant in the fall. Phytoplankton in PWS is exchanged with the adjacent Gulf of Alaska, but imports and exports are assumed to balance each other for the purpose of this model.  Monthly phytoplankton production esti-mates were provided by Dr. C. Peter McRoy (UAF Institute of Marine Sciences, pers. comm.) based on data published by Goering et al. (1973). The months of January, Febru-ary, June, September, and November were missing, so these values were interpolated to calculate the annual primary production in PWS (114 t C⋅km-2; Table 8). This estimate is low, by approximately a factor of two, because observations and measurements were made in Port Valdez/Valdez Arm where glacial silt shades the system (Goer-ing et al. 1973). Thus, 114 t C⋅km-2 converts to an annual produced biomass wet weight of 1,105 t⋅km-2, assuming 0.1 g C = 1 g wet weight (Dalsgaard and Pauly 1997), and doubling this value leads to a biomass pro-duction estimate of 2,210 t⋅km-2. This value must be added to the annual primary produc-tion of macroalgae and eelgrass (501 t⋅km-2; from Dean, this vol.) for a total annual pri-mary production estimate of 2,711 t⋅km-2.  P/B values of 190 and ecotrophic efficiency (EE) values of 0.95 were used to allow the model to calculate phytoplankton biomasses, with Q/B values set at zero, as required for autotrophs. Flagellates and diatom groups were aggregated since they were not ecol-ogically distinguishable with the input pa-rameters available to us and since predation on phytoplankton was always equally split between these groups.   Nearshore Phytoplankton Thomas A. Okey Fisheries Centre, UBC, Canada To ensure dynamic stability of the PWS food web, the phytoplankton group was split into ‘nearshore’ and ‘offshore’ groups, with the 20 m isobath serving as boundary (suggested by S. Pimm, U. of Tenn., pers. comm.; also see Pauly 1998). Diet compositions were allo-cated based on the strata (nearshore vs. off-shore) of each predator of phytoplankton (nearshore or offshore) except for jellies whose phytoplankton consumption proportion (10% of their diet) was allocated by the pro-portion of livable phytoplankton space be-tween the two strata (based on a mean eu-photic zone limit in PWS of 25 m (D. Eslin-ger, UAF Institute of Marine Sciences, pers. comm.), 7.2% of livable phytoplankton space is located nearshore of the 20 m isobath and 92.8% is located in the offshore zone). This proportional allocation would reflect feeding opportunities of jellies on phytoplankton. Other input parameters remained the same as offshore phytoplankton.  ZOOPLANKTON Offshore Zooplankton R. Ted Cooney Institute of Marine Sciences University of Alaska, Fairbanks, USA For the purpose of the PWS model, it seemed reasonable to divide the zooplankton into her-bivorous zooplankton, omnivorous zooplank-ton, and carnivorous jelly plankton. Herbivo-rous zooplankton include copepods, larva-ceans, pteropods, and cladocerans. Omnivo-rous zooplankton include euphausiids, amphi-pods, larval fishes, chaetognaths, and deca-pods. Carnivorous jellies are covered in the following section.  Table 8. Monthly changes in production in PWS (based on Goering et al. 1973). Primary production Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Totals (t⋅km-2⋅year-1) t C⋅km-2 0.2 0.7 1.3 59.6 19.6 10.6 1.6 3.9 5.4 7.0 3.6 0.2 113.8 t ww⋅km-2 2.2 7.4 12.5 596.4 195.6 105.7 15.8 38.5 54.5 70.5 36.4 2.2 1,137.8 16     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      This categorization is not strictly correct, but the approximations should work well for the model. Mesozooplankton (our ‘herbivores’) are all plankters <= 1 mg⋅individual-1. Macrozooplanktors (our ‘omnivores’) are animals > 1mg⋅individual-1. Since some of these animals 'grow through' the mesozoo-plankton category to macrozooplankton at an older age, categories such as juvenile euphausiids and amphipods are thrown in with herbivores when they are young, and with macroplankters when they mature. Zooplankton collections, obtained since the spring of 1994 for project H of the EVOS Trustee Council sponsored Sound Ecosys-tem Assessment (SEA) program, are avail-able for estimates of standing stock from 740 samples collected from 1994-1997. Most of these collections were vertical tows integrating the upper 50 m. A ½-m diameter ring net with 0.33 mm Nitex mesh was used in the field. Some other samples were ob-tained from hatcheries in PWS. These were collected using a ½-m ring net fished verti-cally (by hand) in the upper 20-m (0.25-mm Nitex). All collections were processed in the University of Alaska, Fairbanks Institute of Marine Science (UAF/IMS) plankton labo-ratory using standard subsampling and weighing practices. Numbers and biomass for discrete life stages, species, species composites, genera and more general taxo-nomic categories were recorded. For the analysis supporting the devel-opment of a mass-balance model, monthly averages were determined over all years (1994-1997) for total zooplank-ton (summing across all taxonomic cate-gories), and for specific taxa judged to be important for higher-level consumers in Prince William Sound. Zooplankton densities are highly variable over the course of a year in PWS (Figure 3, and Appendix 3). Wet weight biomass (total, or for specific taxa or size groups) as g⋅m-3, was determined by converting numbers per m3 to g⋅m-3 from average wet weights of in-dividuals. Monthly means were then reported for: (1) total zooplankton (all stages and taxa); (2) for zooplankters <= 1 mg wet weight/individual; (3) for zooplankters > 1 mg/individual; (4) for copepods; (5) for ptero-pods; (6) for amphipods; (7) for larvaceans; (8) for euphausiids; (9) and for a composite (by difference) of everything else (see Appen-dix 3). There were three months for which there were no samples, August, November and January. These means were determined by interpolation.   Monthly estimates (g⋅m-3) were converted to standing stock in kg⋅m-2 over a depth of 300 m (and 100 m; not used here). Prince William Sound has depths to 720 m. The 300 m depth was chosen to generate stock estimates ap-proximating values for the entire water col-umn over the entire Sound for the mass bal-ance modelling. The means of the monthly estimates were then multiplied by the propor-tion of PWS area deeper than 20 m to derive offshore PWS zooplankton biomass densities (beyond the 20m isobath) on a PWS-wide ba-sis (Table 9). The diverse net zooplankton community in PWS is dominated by copepods, some of which produce several generations per year, Figure 3. Seasonal changes in PWS zooplankton, upper 50 m, all years, all locations 050100150200Jan Mar May Jul Sep NovTime (months)Biomass (t/km2 )OthersEuphausiidsLarvaceansAmphipodsPteropodsCopepodsECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     17 while others reach up to a year of age. Pseu-docalanus spp. are the most common small copepods in our samples and produce sev-eral broods per year. The larger copepods are dominated by Neocalanus spp. and Ca-lanus marshallae. These calanoids produce one (Neocalanus), at most two (Calanus) generations per year. Pteropods, Clione Limacina and Limacina helidina are numer-ous, as are euphausiids (Thysanoessa spp. and Euphausia pacifica). The Pteropods re-produce continuously after the spring bloom, which starts in April; euphausiids in the re-gion are believed to live for 2-3 years. Am-phipods were judged to exhibit life histories similar to the euphausiids, and the larvacean production cycle more like that of the ptero-pods. The ‘other zooplankton’ consists of a variety of jelly forms (hydromedusae and ctenophores), larval fishes and meroplank-ton. There is a huge range in production/biomass (P/B) ratios in the literature (see Tranter 1976; Valiela 1995). The values of P/B se-lected here (5 year-1 for herbivorous zoo-plankton and 15 year-1 for omnivorous zoo-plankton) fall within the ranges of those re-ported for mixed zooplankton, and for indi-vidual groups or species (overall zooplank-ton, 10 year-1; small zooplankton, 15 year-1; large zooplankton, 5 year-1; copepods , 8 year-1; pteropods, 3 year-1; amphipods, 2 year-1; larvaceans, 3 year-1; euphausiids, 2 year-1; other zooplankters, 2 year-1). These annual estimates were then distributed across the months generally in proportion to the growth cycle of the zooplankters as ob-served in their seasonal signal; lower during the winter months, and higher in the late spring and early summer (Appendix 3 and Figure 3). The amounts of food ingested for each of the groups was determined by applying a gross growth efficiency of 30% to the es-timated monthly production values. These consumption estimates (Q) were then divided by the biomass on the different months to provide the Q/B ratios (50 year-1 for her-bivorous zooplankton and 17 year-1 for omnivorous zooplankton). Parsons et al. (1988) list gross growth efficiencies for zoo-plankton. While the range is quite high, most fall within 20-40%. Harrison et al. (1993) list a Q/B value for herbivorous zooplankton in the Strait of Georgia of 10.5 year-1. The calcu-lated values obtained using a 30% gross growth efficiency range from 7-50 year-1 (see Appendix 3).  In the absence of a way to measure or calcu-late ecotrophic efficiency (EE), i.e., the frac-tion of the production consumed or exported, the default value used by Dalsgaard and Pauly (1997) for a preliminary mass balance model of Prince William Sound is suggested. While this value may be a reasonable annual aver-age, it probably does not apply well to monthly production values. Obvious increases in zooplankton during the late spring and summer months suggest an uncoupling of the system for some of the taxa examined in this study. Appendix 3 lists monthly estimates of EE, generally phased inversely with levels of per capita production. However, the fraction consumed (EE) is not allowed to decline be-low 0.9 in any month. Levels of Uncertainty The information provided here is probably most accurate at the level of monthly biomass in the upper 50 m. Extending this value to greater depths implies an unknown bias. Cer-tainly, zooplankton live at all depths in the region, but at least during the spring and summer, the biomass of most populations is greatest near the surface. I presume that by distributing the upper 50 m derived values of g⋅m-3 to a deeper water column there will be a depth below which the overall estimate will become significantly positively biased. On the other hand, the small size of the net used by SEA probably under-represented some of the taxa (euphausiids and amphipods). This nega-tive bias will be partially corrected when sur-Table 9. Annualized Ecopath parameters for zooplankton in PWS. Zooplankton category Biomass (t⋅km-2) P/B (year-1) Q/B (year-1) EE Herbivorous zooplankton 30.0 15 50 ≥ 0.90 Omnivorous zooplankton 15.4 5 17 ≥ 0.90 18     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      face values are extended over a deeper col-umn. Damkaerr (1977), based on 1-m ring net, reported more than 1 kg⋅m-2 in tows from the deepest part of the Sound to the surface, or roughly an order of magnitude greater than reported here for ‘total zoo-plankton’ summed over 300 m in June. This suggests extreme interannual or spatial vari-ability, or very strong negative bias associ-ated with the smaller net used in this study, or something else. Because the amount in-gested to account for the calculated zoo-plankton production is considerably higher than the estimate of annual primary produc-tion, I suspect the values of zooplankton standing stock are too high. When the upper 50 m values are distributed over 100 m, the zooplankton ingestion is more in line with primary productivity, particularly since the latter (presented by McRoy, this volume) seems too low by about a factor of 2 (Goer-ing et al. 1973). The production to biomass (P/B) ratios used here to predict monthly production from monthly standing stock estimates are arbi-trarily chosen from a small range of contro-versial literature values. I suspect that they are accurate only in the most general sense, and may generally be considered conserva-tive.  Ingestion per unit biomass (Q/B) values are calculated from a single value of gross growth efficiency (30%), an average for several zooplankters reported in the litera-ture (range 7-70% year-1). For the larger zooplankton like euphausiids, this growth efficiency may over-estimate food consump-tion. This approach may introduce a slight positive bias to estimates of food consumed monthly by zooplankton.  Nearshore Zooplankton Robert J. Foy Institute of Marine Sciences University of Alaska, Fairbanks, USA The ‘nearshore zooplankton’ group in the PWS model consists of omnivorous zooplank-ton and herbivorous zooplankton collected within the 20 m depth contour. Zooplankton samples were collected from May, 1996 to March, 1997 with 300 µm mesh net vertical tows from the head of four bays in PWS as part of the EVOS Trustee Council sponsored Sound Ecosystem Assessment (SEA) pro-gram. Biomass estimates determined from subsample counts and individual weights were pooled for monthly means. Biomass data in April and September were interpolated be-cause no samples were collected. The means of monthly nearshore zooplankton biomass values were expressed on a Sound wide basis by multiplying by the proportion of PWS area in the nearshore zone (0.1586). Summary data are presented in Table 10 and monthly data are presented in Appendix (3). Production of zooplankton is highly variable and difficult to estimate (Lalli and Parsons 1993; Valiela 1984). P/B estimates for om-nivorous and herbivorous zooplankton were derived from estimated annual P/B values for euphausiids (annual P/B=7.9 year-1) and cope-pods (annual P/B=27.0 year-1), respectively, off southwestern Vancouver Island, British Columbia (Robinson and Ware 1994). Monthly estimates were adjusted in proportion to the biomass data.  Food consumption (Q) for each nearshore plankton group was calculated by assuming that the monthly production value was based on a 30 % growth efficiency, which falls within the range of other aquatic invertebrates (10-40 %) (Parsons et al. 1984). Gross growth efficiency can be seasonally variable depend-ent on food concentration and temperature (Raymont 1983). Annual Q/B values ranged from 26 to 90 year-1. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     19 Diet composition was set at the same levels as the offshore zooplankton groups, dis-cussed in the previous section. Predation by other groups on nearshore zooplankton was allocated from the offshore zooplankton group in proportion to the water volume in the two areas; for each zooplankton group, 0.6% of its zooplankton proportion was al-located to nearshore zooplankton while the remaining (99.4%) remained in the offshore portion, beyond the 20 m isobath, unless predator distribution information indicated spatial heterogeneity of feeding by particular predators. These percentages correspond to 14.2 km3 nearshore of the 20 m isobath (as-suming a mean depth of 10 m) and 2200 km3 offshore (integrated to 300 m).  Carnivorous Jellies Thomas A. Okey Fisheries Centre, UBC, Canada Robert J. Foy Institute of Marine Sciences University of Alaska, Fairbanks, USA Jennifer Purcell Horn Point Lab, U. of Maryland  Center for Env. Sci., USA The biomass density estimate of PWS car-nivorous jellies (6.39 t⋅km-2) was estimated using two different data sets, one for jellies in PWS open water areas (J. Purcell, unpub-lished data) and one for jellies in PWS near-shore areas (R. Foy, unpublished data). This estimate is a weighted mean of the two in-dependently derived estimates, based on the proportion of total area that each zone repre-sents.  The biomass of carnivorous jellies in the offshore surface waters of PWS was esti-mated to be 7.065 t⋅km-2. This value was derived by multiplying the volumetric biomass estimate, 94.2 t⋅km-3 (J. Purcell, unpublished data), by 0.3 km, the aver-age depth of PWS, and dividing by four to account for assumed declines in jelly densities with increasing depth.  The biomass density of nearshore jellies was estimated to be 2.79 t⋅km-2. This estimate was derived by multiplying the nearshore volumet-ric biomass estimate (278.9 t⋅km-3; R. Foy, unpublished data) by 0.01 km, the assumed average depth inshore of the 20 m isobath. Production to biomass ratios (P/B) for car-nivorous jellies were derived from maximum daily P/B reported for gelatinous predators in Saanich Inlet, British Columbia (Larson 1986). A maximum daily P/B of 0.1 year-1 was assumed for June (when maximum pro-duction occurs) while the other months were calculated proportionally to biomass. The an-nual P/B would then be 8.82 year-1, which is consistent with that of British Columbia ge-latinous predators (P/B = 5-10 year-1). The annual Q/B was set at 29.4 year-1, and the diet composition used for carnivorous jellies was 67% herbivorous zooplankton, 23% omnivo-rous zooplankton, and 10% phytoplankton.  BENTHIC INVERTEBRATES Shallow Large Infauna Tom A. Dean Coastal Resources Associates Vista, California, USA These are larger (generally greater than 20 mm) infauna found at depths less than 20 m. Clams make up the majority of larger infaunal biomass in the nearshore. These include Pro-tothaca staminea, Saxidomus giganteus, Cli-nocardium spp., Macoma spp. and others. Surveys of subtidal and intertidal clam densi-ties were conducted in Herring Bay, Bay of Isles, and Montague Island portions of the Sound in 1996 and 1997 by G. Van Blaricom, A. Fukayama, S. Jewett, and T. Dean (Hol-land-Bartels et al. 1997 and unpublished data). Sampling was conducted from the intertidal to Table 10. Annual summary data for nearshore zooplankton Group Biomassa (t⋅km-2) P/Bb (year-1) Q/Bb (year-1) Herbivorous zooplankton 0.097 27.0 90.0 Omnivorous zooplankton 0.079 7.9 26.3 a. Means of monthly nearshore biomass values expressed on a PWS-wide basis (times 0.1586; see Appendix 3). b. Sums of monthly P/B or monthly Q/B values (see Appendix 3) 20     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      depths of 15 m using either shovels (in the intertidal) or a suction dredge (in the subtidal). We assume here that these data are representative of the entire Sound, and are comparable to pre-spill (1980-1989) density estimates. The estimated clam biomass, for both intertidal and subtidal habitats, was ap-proximately 80 t ww km-2. On a sound wide basis, this is estimated at 12.5 t km-2. There are no data on seasonality, but we suspect that there is no appreciable change in standing stock biomass with season. The dominant clams are primarily sus-pension and deposit feeders, and we as-sume that about 50% of the diet is phy-toplankton, and 50% detritus. Feder and Jewett (1986, Table 12-9) esti-mated that the infaunal biomass at Hinchen-brook entrance was 343 g m-2, and this pro-duced 4.6 g C⋅m-2 ⋅year-1, corresponding to 222 g⋅ww⋅m-2 ⋅year-1. These estimates were for depths greater than 20 m, from an ex-tremely productive portion of the Sound, and are thus probably not representative of Sound wide conditions. However, we use these data to provide an estimate of a P/B ratio for shallow large infauna, of about of 0.6 year-1. The Q/B ratio for large infauna is estimated at 23 year-1, based on estimates given by Guénette (1996) for the North Pa-cific. Some clams (especially Protothaca) are harvested. A catch of 0.003 t·km-2·year-1 for PWS clams is given in Trowbridge (1996).   Shallow Small Epifauna Tom A. Dean Coastal Resources Associates Vista, California, USA Shallow small epibenthos are defined as non-motile or slightly motile invertebrates of less than 5 cm in size living on or near the bottom at depths less than 20 m. These are generally found on hard substrates or at-tached to algae/eelgrass. Small epifaunal or-ganisms in the nearshore zone include a vari-ety of invertebrate taxa. Dominant forms in-clude barnacles, littorine and lacunid snails, mussels, limpets, chitons, and amphipods, small crabs, and other snails and crustaceans. Highsmith et al. (1994) estimated the biomass of intertidal epifauna within several depth strata and habitat types in PWS in 1990 and 1991 (Table 11). The average biomass at uno-iled sites was 701 t km-2 (Table 12). On a Sound wide basis, accounting for the whole area, this is equivalent to 23 t⋅km-2. Estimates for epifaunal invertebrate abun-dance in the nearshore subtidal zone were made within three areas (Herring Bay, Bay of Table 11. Biomass (t ww km-2) of small intertidal epifauna in different strata and habitat types (from Highsmith et al. 1994). Depth strata and habitats are as defined in Tables 2 and 4. (from Highsmith et al. 1994). Habitat type Depth stratum May 1990 August 1990 May 1991 Mean High 271 263 362 Mid 493 553 1,173 Low 397 395 411 Sheltered rocky Mean 387 404 649 480 High 209 117 181 Mid 543 573 424 Low 328 617 406 Coarse textured Mean 360 436 337 378 High 3,454 13,456 197 Mid 6,440 8,354 95 Low 2,815 6,887 91 Estuarine Mean 4,236 9,566 128 4,643 High 149 66 540 Mid 1,324 745 854 Low 1,128 181 1,671 Exposed rocky Mean 867 331 1,022 740 Table 12. Weighted mean biomass (t ww⋅km-2) of small epibenthic invertebrates in the intertidal zone, Prince Wil-liam Sound (from Highsmith et al. 1994). Habitat Type Biomass density (t⋅km2) % of area Biomass in PWS Sheltered rocky 480 30 144 Coarse textured 378 41 154.98 Estuarine 4,643 5 232.15 Exposed Rocky 740 23 170.2 Fine Textured 0 1 0 All 6,241 Total 701.33 ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     21Isles, and Northern Montague Island) by T.A. Dean and S. C. Jewett in 1997 (unpub-lished data). Airlift samples of invertebrates were collected at systematically selected sites at depths of 1 to 3 m in order to esti-mate the abundance of food available to har-lequin ducks. Several of the dominant taxa, including littorine and lacunid snails, lim-pets, chitons, amphipods, and other snails and crustacea were counted and weighed. The average biomass was 7.5 t⋅ww⋅km-2. This is likely an underestimate of biomass for the Sound as a whole, because densities of invertebrates tend to be higher at more exposed sites that were not sampled, and several common invertebrate taxa (e.g., ser-pulid polychaetes) were not sampled. How-ever, if we assume that the biomass is repre-sentative of the entire subtidal zone in PWS, then the average biomass, on a Sound wide basis, is estimated at 1.2 t⋅km-2. Thus, the total biomass of small epifauna in both inter-tidal and subtidal habitats is approximately 8.7 t⋅km-2. There are no good quantitative data on sea-sonality of biomass for small epifaunal in-vertebrates in the nearshore. However, data in Highsmith et al. (1995) suggest that inter-tidal biomass peaks in early summer (June) following spring recruitment. We suspect that minimum biomasses occur in late March. This is just after the breakup of ice formations that cause significant mortality to intertidal organisms, and just prior to the Spring recruitment and phytoplankton blooms. We also suspect that there is less seasonality in the subtidal than the intertidal assemblages. We assume here that the minimum biomass occurs in March and is about 75% of the peak in June. The trophic levels and diets of small epifauna in the nearshore are extremely varied. We esti-mate that roughly 75% of the biomass consists of barnacles, mussels, and other filter feeders that feed primarily on phyto-plankton and detritus. Most of the remaining biomass consists of grazers (e.g. littorines, chitons, and limpets) that feed primarily on smaller algae. We esti-mate that the composite diet consists of about 40% detritus, 35% phytoplankton, 20% algae, and 5% small nearshore epifaunal inverte-brates. The P/B ratio for small epifauna is assumed to be about 2 year-1, equivalent to that for larger epifauna as described by Feder and Jewett (1986). The Q/B ratio is assumed to be ap-proximately 10 year-1, based on values given in Guénette (1996) for epifauna.  Shallow Large Epifauna Tom A. Dean Coastal Resources Associates Vista, California, USA Shallow large epifauna are defined as gener-ally motile invertebrates that are greater than 5 cm and live on or near the bottom from the intertidal to depths of 20 m. These are mostly sea stars (Pycnopodia helianthoides, Dermas-terias imbricata, Evasterias troschelii, etc.) and crabs (mostly Telmessus cheriagonus). Dean et al. (1996b) surveyed large epibenthic invertebrates in 4 habitats in Western Prince William Sound in 1990. Based on data from unoiled control sites, we estimate the average density of large epibenthic invertebrates was 0.27 individuals⋅m-2 (Table 13). These include mostly starfish (mainly Dermasterias and Pycnopodia) and crabs (mainly Telmessus). An average Dermasterias has a wet weight of about 75g (T. Dean, unpublished data). Based on the assumption that all large epifauna are of about the same weight, this leads to a bio-Table 13. Density and biomass of large epibenthic invertebrates in 0 to 20 m depths in Prince William Sound, Alaska (from Dean et al. 1995b and Jewett et al. 1995; extrapolated from (individuals⋅100 m-2) and (g⋅ww⋅m-2) respectively). Taxa Density (ind.⋅km-2) Biomass (t⋅km-2) Pycnopodia helianthoides (adults) 1,000 7.5 Dermasterias imbricata (adults) 1,000 7.5 Evasterias troschelii 200 1.5 Telmessus cheriagonus 300 2.3 Others 200 1.5 Total 2,700 20.0 22     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      mass density of roughly 20 g ww⋅m-2. On a Sound wide basis, the standing stock bio-mass is estimated at 3.1 t⋅km-2. The diets of Pycnopodia (Holland-Bartels et al. 1997) and Evasterias (O’Clair and Rice 1985) consist mostly of clams and snails. Dermasterias eat a wider variety of benthic invertebrates (Rosenthal et al. 1974). Tel-messus consume eelgrass and associated small epifauna (McConnaughey and McRoy, 1979). It is estimated that the aver-age diet of large epibenthic invertebrates consists of 80% shallow small epifauna, 19% nearshore small infaunal invertebrates, and 1% eelgrass. There is no appreciable harvest of nearshore epibenthic invertebrates in PWS. Large epibenthic invertebrates that have been of historical commercial value (crabs and shrimp) are generally restricted to depths greater than 20 m. The P/B ratio for large epifauna is assumed to be about 2 year-1 based on estimates of Feder and Jewett (1986). The Q/B ratio is assumed to be approximately 10 year-1, based on values given in Guénette (1996) for epifauna.  Small Infauna Stephen Jewett Institute of Marine Sciences University of Alaska, Fairbanks, USA Information on the macrobenthos in Prince William Sound are mainly available from the pre-EVOS investigations of Hoskin (1977), Hoberg (1986) and Feder and Jewett (1988) and post-EVOS studies of Feder (1995), Jewett et al. (1995) and Jewett and Dean (1997). The last two investigations, which were conducted in 1990, 1991, 1993 and 1995, targeted depths < 20 m; the others targeted depths > 20 m. Information in-cluded in this synopsis is from pre- and post-spill sources. All sampling generally oc-curred between April and August. Macro-benthos in these investigations refer to all benthic invertebrates larger than 1 mm that were sampled with a 0.1 m2 van Veen grab or suction dredge. These mainly include infaunal organisms that live within the top 10 cm of substrate and small, slow-moving or sessile epifauna.  Estimates of proportions of subtidal habitats at < 20 m, 20-100 m, and > 100 m in PWS are presented in Table 14. The extent of coverage of < 20 m depth habitats of bays predominated by kelps (Laminaria and Agarum ) and eel-grass (Zostera ) was estimated using side-scan sonar and systematic surveys by divers along segments of the western portion of the Sound (Jewett and Dean, unpubl.). The extent of hard (unsampleable) and soft (sampleable) sub-strates at depth > 20 m was estimated by S.C. Jewett (pers. obs.) and T.A. Dean (Coastal Resources Associates, pers. comm.).  Macrobenthic biomass estimates from differ-ent habitats in the Sound are presented in Table 15. Estimates from < 20 m depths are mainly from relatively unexposed bays in the Knight Island vicinity where kelps and eel-grass predominate (Jewett et al. 1995; Jewett and Dean 1997). Estimates from 20-100 m depths are from western PWS, mainly in the Table 14. Proportion of substrate types in different PWS depth zones (%). Depth (m) Area <20 20-100 >100 Hard substrate 76.0 30 10   Laminaria/Agarum bays 37.5 ---- ----   Unsampleable 38.5 ---- ---- Soft substrate 24.0 70 90   Zostera bays 7.5 ---- ----   No vegetation 16.5 ---- ---- Table 15. Estimates of macrobenthic biomass in PWS.  Area Biomass (t⋅km-2) Weighting factor Biomass (t⋅km-2) < 20 m 83.8±12.1 0.615 51.5   Laminaria/Agarum 76.0±19.9 0.375 28.5   Zostera bays 87.9±17.7 0.075 6.6   Other 81.9 0.165 13.5 20-100 m 88.1±16.6 0.700 61.7 > 100 m 19.1±3.5 0.900 17.2 Combined > 20 m  -- 24.7a a. calculated by adding the products of the biomasses and areal proportions of the two preceding groups. Proportions in Table 14 and Table 8 were used for this calculation. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     23 Knight Island/northern Montague Island vi-cinity (Feder, 1995; Table 4; Hoberg, 1986; Append. 7). Estimates from > 100 m depths are from three fjords in western PWS (Der-ickson and McClure Bays and Blue Fjord; Hoskin, 1977) and northern PWS (Port Val-dez; Feder and Jewett, 1988; Table 2) and non-fjords mainly in the Knight Is-land/northern Montague Island vicinity (Feder, 1995; Table 4). Not included in the > 100 m stratum estimate is the high biomass of 417 t km-2 in the region of Hinchinbrook Entrance (Feder and Jewett, 1986; Table 12-6). This region, which is very dynamic, ap-pears to be exceptionally productive in com-parison to other areas > 100 m. I assumed that there is little seasonal biomass variabil-ity.  The dominant faunal groups (% biomass) by depth strata are presented in Table 17. Val-ues for < 20 m are estimated from Jewett et al. (1995; Append. P and T). Values for depths 20-100 m and  > 100 m are from Hoberg (1986) and Hoskin (1977), respec-tively. Other sources of information on dominant groups at depths > 20 m report dominance in terms of numerical abundance rather than biomass (Feder and Jewett 1988; Feder 1995). Estimates of P/B, Q/B and diet matrix of macrobenthos are presented in Table 16. P/B values for depths < 100 m ( 0.6) are estimated using previously estimated P/B values for specific taxa from Feder et al. (1989; Appen-dix 3). P/B values for depths > 100 m (1.4) are from Feder and Jewett (1988; Table 3). The Q/B value used here, 32 year-1, is the same used for infaunal macrofauna by Guénette (1996). I assumed that the phytoplankton con-sumed by suspension feeders is composed of one-half diatoms and one-half flagellates.  Deep benthic groups and meiofauna Thomas A. Okey Fisheries Centre, UBC, Canada Deep large infauna These are larger infauna (generally greater than 20 mm), mostly clams, found at depths greater than 20 m in PWS. Clams have been sampled at depths deeper than 20 m in PWS (Feder and Blanchard 1998), but I could find no data that can be used to reliably estimate the biomass density of deep large infauna in PWS. Assuming that clam biomass density below 20 m is generally one fourth that of the shallow zone, the density value of 80 t ww⋅km-2 provided by Dean (this vol.) for shal-low large infauna was divided by four then expressed on a sound wide basis by multiply-Table 16. Estimates of P/B, Q/B and diet matrix (% abundance) of macrobenthos in PWS. Depth zones (m) P/B (year-1) Q/B (year-1) Detritivores (Detritus) Suspension feeders (Phytoplankton) Predators/ scavengers (Cannibalism) < 20 0.6 23 0.25 0.60 0.15 20-100 0.6 23 0.77 0.20 0.03 > 100 1.4 23 0.80 0.15 0.05 > 20a 0.94 23 0.78 0.18 0.04 a)  >20 m values were calculated by multiplying the values in the preceding two categories with the corre-sponding biomass proportions in those two depth zones then summing the respective products; biomass propor-tions were calculated from Table 11. Table 17 Dominant macrobenthic groups in PWS (% biomass). Depth zones (m) Anthozoa Polychaete Bryozoan Bivalve Gastropod Echiuran Holothurian Ophiuroid < 20 <1 11 4 74 5 <1 <1 1 > 20a 4 26 <1 26 <1 3 12 8 20-100 6 <1 <1 28 <1 5 20 13 > 100 <1 60 <1 23 <1 <1 <1 <1 a) Figures were calculated by multiplying the values in the following two categories with the corresponding biomass proportions in those two depth zones, then adding the respective products; biomass proportions were calculated from Table 11. 24     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      ing by the proportion (0.81) of deep zone area in PWS from Table 1. The deep large infauna biomass estimate is 16.2 t⋅km-2.  The numerically dominant clams are primar-ily deposit feeders in the deep zone, and we assume that about 10% of the diet is phyto-plankton, and 90% detritus. P/B ratio for shallow large infauna, of 0.6 year-1, is also used for deep large infauna. The Q/B ratio for large infauna is estimated at 23, based on estimates given by Guénette (1996) for the North Pacific. Deep Epifauna This group is made up of both motile and non-mobile invertebrates living on, but not in, the sea floor at depths greater than 20 m. There is inadequate information to reliably estimate biomass of PWS epifauna in the zone deeper than 20 m for any period after the EVOS. Epifaunal biomass has been es-timated for one location in PWS prior to EVOS, but information compiled in Feder and Jewett (1986) reveals that the composi-tion of epifaunal assemblages is highly vari-able among PWS sites (Table 18). The de-gree of temporal variability of deep epifauna is even less known, though it might be considerable, as indicated by recent fluctuations (declines) in crabs and shrimps throughout Alaska (NMFS 1996). A rough biomass estimate of 1.5 t⋅km-2 was derived by giving equal consideration to the pre-spill biomass value for Port Etches and a pre-spill value derived for the lower Cook Inlet, a nearby setting similar to the central area of PWS (Table 18). The mean of these two values, 1.85 t⋅km-2, was multiplied by 0.81, the proportion of PWS area deeper than 20 m. A biomass estimate derived from 1975 trawls on the shelf adjacent to PWS was not used in the derivation of the rough estimate, but it is included in (Table 18). Calculated values for Tanner Crab and shrimps are also included in the table for comparison.  The P/B ratio for epifauna is assumed to be about 2 year-1, based on estimates given by Feder and Jewett (1986). The Q/B ratio is assumed to be approximately 10 year-1, based on values given in Guénette (1996) for epi-fauna. A catch rate of 0.143 t·km-2·year-1 for PWS epifauna is provided in Trowbridge (1996). Included in this rate is pink and other shrimps, king crab (red, blue, brown), and tanner crab. Meiofauna ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     25The P/B for meiofauna was set at 4.5 and the Q/B was set at 22.5 (Tom Shirley, UAF In-stitute of Marine Sciences, pers. comm.). The biomass was estimated by Ecopath with the ecotrophic efficiency set at 0.95.  PLANKTIVOROUS ‘FORAGE FISHES’ Salmon Fry Thomas C. Kline, Jr. Prince William Sound Science Center Cordova, Alaska, USA  Salmon fry consist of five species of On-corhynchus. The relative contribution of these five species in decreasing abundances is pink (O. gorbuscha) > chum (O. keta) > red (O. nerka) >silver (O. kisutch) > king (O. tshawytscha). One hundred per cent of the last two species are artificially propa-gated (by hatcheries) in PWS. In this section, salmon fry are divided up between those < 6 cm and those ≥ 6 cm in length. Those < 6 cm enter the system either from natural habitats  (spawning redds on beaches or natal streams) or hatcheries. Those > 6 cm leave PWS at about 11 cm. The ration-ale for division of salmon fry into two func-tional groups is based upon the SEA project findings that suggests a predation size refuge near 6 cm (Willette et al. 1996). Mortality rate of salmon is highest upon entry into the ma-rine system (Parker 1968). Furthermore, mor-tality nearly compensates biomass increase for fry < 6 cm while average biomass of fry > 6-cm doubles. Splitting into the two functional groups thus also leads to more reasonable means, then to compute a grand mean for ju-veniles, as required for the single juvenile group, which Ecopath can accommodate for each species. Approximately 500 million salmon fry enter PWS from hatcheries each year while 300 million enter from natural stocks. The average entry date into PWS from these sources is May 1 (PWSFERPG 1993), and the average weight is 0.25 g  (PWSFERPG 1993). This number multiplied by the starting population leads to a biomass estimate of 160 t. There is 32% survival after the 40 day period when the fry are less than 6 cm. Thus, the ending popu-lation size is 256 million. This population size Table 18. Estimated biomass of benthic epifauna > 20 m depth, PWS (from Feder and Jewett 1986). Area (and depth) (m) Biomass (t⋅km-2) Species (group) (% weight) Port Etchesa  (85-150) 0.8 Sunflower star (62), Pink shrimp (28), Tanner crab (4), Mollusks (0.2), Other  (5.8) Rocky Bayb  (30-100) --- Echinoderms (87), Crustaceans (5.9), Molluscs (3.2) Zaikof Bay (20-100) --- Echinoderms (50), Crustaceans (45) Outer Simpson Bay (30-50) --- Sea pens, cockles, brachiopods, basket stars Outer Port Gravina (50-130) --- mud stars, pink shrimp Outer Port Fidalgo (90-170) --- crinoids, basket stars Inner Galena Bay (30-130) --- feather star, shrimps Columbia Bay (150-275) --- sea pens, mud stars Unaquik Inlet (175-212) --- mud stars, pandalid shrimps Port Wells (275-400) --- mud stars, heart urchins, cucumbers NEGoA shelfc 2.1 --- Lower Cook Inlet (>25) d 2.9 --- PWS Tanner Crab, 1989 e 0.24 --- PWS Shrimp, 1989 e  0.01 --- Mean of selected estimatesf 1.5 --- a) Feder and Hoberg (1981) in Feder and Jewett (1986);  b) Feder and Hoberg (1981) and Hoberg (1986) in Feder and Jewett (1986); c) Feder and Matheke (1980) in Feder and Jewett (1986); d) Feder and Paul (1981) in Feder and Jewett (1986); e) NMFS (1993); f) mean of lower Cook Inlet and Port Etches.26     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      when multiplied by an average weight of 1.23 g (PWSFERPG 1993) results in a bio-mass of 315 t. Thus, the average biomass of salmon fry < 6 cm is 238 t. The mortality estimated for the second period, 70 days, reduces the population to 84,480,000 using a mortality rate of 0.0158 day-1. The popula-tion biomass, using an average weight of 10.04 g (PWSFERPG 1993), at this time is thus 848 t. This biomass is exported from PWS. The average biomass for fry in PWS > 6 cm is 582 t. Since the import was 160 t, the net export is 422 t. The area of PWS used for the areal density calculations is 9,059 km2. The P/B ratio for both salmon fry groups can be approximated on the assumption of linear growth and mortalities, which holds for short intervals. Thus, for a short period ∆t = t2-t1,   BWNBBP )( ⋅∆+∆≈   where: N∆ = change in number of individuals (N1 – N2); W = mean weight of individual ((W1+W2)/2); B = biomass (N ⋅ W ); B∆ = change in biomass (B2 – B1); B = mean biomass (B1 + B2)/2.  The resulting P/B value was then expressed on an annual basis for the model. Using this approach and the preceeding estimates, P/B for salmon fry <6 cm is calculated to be 18.542 year-1 (2.032 ⋅ (365 days/40 days)). The P/B for salmon fry >6 cm, during their 70 day residence time in PWS, is cal-culated to be 8.666 year-1  (1.662 ⋅ (365 days/70 days)). The biomass-weighted average of the two values (9.844 year-1) is the P/B of the aggregated group (Table 20 and Table 21).  Salmon fry consumption rate was esti-mated based on smaller sized fry for which data are available. Consumption rate ranged from 4.5 to 31.5 percent body weight per day (Table 1 in Willette et al. 1996) for an average of 17.2 g. This corresponds to an estimate of Q/B = 62.8 year-1. The range on this value is 16.4 to 115 year-1. There is a net export of fry < 6 cm  (see, above). The import, i.e., negative export: -160 t, or -0.018 t km-2 . The export is 256 t, or 0.029 t km-2. Net export is thus (256-160) t or 96 t, i.e., 0.011 t km-2. There is also a net export of fry > 6 cm. The import, i.e., negative export is -256 t, or -0.029 t km-2 . The export is 665 t, or 0.076 t km-2. Net export is thus (665- 256) t, or 409 t. In areal units: 0.046 t km-2. Fry smaller than 6 cm are present in PWS for 30 days in May and ten days in June. Fry greater than 6 cm are present for 20 days in June, 30 days in July, and 20 days in August. Salmon fry are absent for all other months. Table 19. Diet composition (in %) of pink and chum salmon fry in PWS (1994-1996)a Prey Categories Pinks Chums Meansb Small pelagic fishes 36.5 46.4 37.7 Herbivorous zoopl. 30.3 20.6 29.1 Omnivorous zoopl. 15.6 22.2 16.4 Shal. sm. epibenthos 16.4 10.6 15.7 a)  Adapted from APEX-SEA data provided by M. Stur-devant (NMFS Alaska Fisheries Science Center); b)  Proportionally weighted means of Pink (0.878) and Chum (0.122) salmon. Table 20. Ecopath parameters for salmon fry > 6cm  (P/B values not used in model parameterization). B Period B (t⋅km-2) min max P/B (year-1) Export (t⋅km-2⋅year-1) Jun 0.035 0.029 0.042 -- -0.029 Jul 0.052 0.042 0.062 -- 0 Aug 0.069 0.062 0.076 -- 0.076 Mean 0.050 0.027 0.073 8.666 0.034 Aggregate(fry 0-12) 0.072 -- -- 9.844 0.045 ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     27 Five-eighths of the salmon fry in PWS origi-nate from hatcheries which biases their distribution into a corridor running from Port Valdez to the central north region as far as Esther Island then running south through Knight Island Pass. and around the east side of Knight Island. From here they are distrib-uted southward and out of PWS into the Gulf of Alaska via the southwest passages (Figure 1). Within this area salmon orient themselves to land staying within the 20 m isobath. It is estimated that about 80% of the fry in PWS are found within this area while the balance are distributed more or less evenly in the remaining area, but within the 20 m isobath. Diet compositions for pink and chum salmon fry in PWS from 1994-1996 are adapted from APEX-SEA data provided by M. Sturdevant (Table 19, Appendix 4). Since these two species make up 96% of the salmon in PWS, proportionally weighted averages of their diets represent a general-ized salmon diet in PWS.  Adult Pacific Herring Thomas A. Okey  and Johanne Dalsgaard Fisheries Centre, UBC, Canada Pacific herring, Clupea pallasi, are school-ing zooplanktivores that are usually found near the surface, but can occur at various depths and may disperse at night. Adults are defined as equal or greater than 18 cm in length; three year olds are considered adults by length and age distribution, but some do not mature until their fourth year. Herring can live up to 19 years in Alaska (Love 1996). Most spawning occurs along the north shore of Montague Island at 0-15 m, but the con-fined spawn at Montague is only a shadow of its former self.  Over half the spawn was formerly in the northeast and north shore of PWS, and the population may now be rebuilding there, and on the east side (mainly Orca Inlet and Sheep Bay) of PWS (E. Brown, UAF Institute of Marine Sciences, pers. comm.). Large summer concentrations of adult herring have been found in SW passages, Esther Passage, Wells Bay and the outer (eastern) coast of Monta-gue.  Adults are widely distributed in the upper 50-100 m, but not as widely as juveniles. Spring and mid-winter distributions are known, but not summer and early fall distributions. They may range offshore to shelf edge of the gulf of Alaska and beyond in summer for food and return in the fall. Substrate for spawning is kelp (over rocky bottom) and eelgrass (over sandy bottom) (E. Brown, R. Foy, and J. Wil-cock, UAF Institute of Marine Sciences and Alaska Dept. of Fish and Game, pers. comm.). Estimates of pre-fishery run herring biomass sharply declined in PWS four years after the EVOS (Figure 4; data include ages 3 and above).  The adult herring biomass estimate of 23,143 t in PWS, or 2.555 t⋅km-2, is the mean of the estimates from the years 1994-1996 (Table 22). The three-year period of the model oc-curred after the sharp herring decline of 1993. biomass estimates were based on Age-Structured Assessment modelling (data pro-vided by J. Wilcock, Alaska Dept. of Fish and Table 21. Ecopath parameters for salmon fry < 6cm  (P/B values not used in model parameterization). B  B Period (t⋅km-2) min max P/Ba (year-1) Export (t⋅km-2⋅year-1) May 0.020 0.018 0.021 -- -0.018 Jun 0.025 0.021 0.029 -- 0.029 Mean 0.022 0.018 0.027 18.542 0.011 Table 22. Estimated Pre-Fishery Herring Run biomass in PWS (1994-1996) a. Year Pre-fishery run biomass (t) Biomass density (t⋅km-2) 1994 19,121 2.111 1995 23,933 2.642 1996 26,376 2.912 Mean 23,143 2.555 a) Data provided by J. Wilcock, Alaska Dept of Fish and Game, pers. comm.) 28     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Game, pers. comm.). The three fisheries sub-sectors that normally catch herring in PWS are spawn on kelp, sac roe by seine and gillnet, and food-and-bait by purse seine or trawl. These fisheries were closed in PWS during the three-year model-ling period (1994-1996) except for a food-and-bait catch of 847 t in 1996 for a three-year mean of 282 t, or 0.031 t⋅km-2 (Morstad et al. 1997). Given that, under equilibrium, F = catch/biomass, a fishing mortality (F) of 0.01 year-1 can be estimated from the above figures. Natural mortality (M) was estimated as 0.53 year-1, as the means of age-specific estimates (age 3 to 8) for herring in the Gulf of Alaska (Wespestad and Fried 1983). Since P/B, under equilibrium, equals total mortality (Z; Allen 1971), and Z = F + M, the P/B ratio for herring can be estimated as 0.53 + 0.01 = 0.54 year-1. The value of Q/B used here, of 18 year-1, is the same as that used for small pelagics (mainly herring) in the Strait of Georgia (Venier 1996a). The diet composition of adult herring was derived from 1994-1996 APEX-SEA data, provided by M. Stur-devant Table 24, Appendix 4).  The estimate for the annual fishery removal rate from 1994-1996 is 2.55 t⋅km-2⋅year-1. It was necessary to increase the Pacific herring P/B ratio from 0.54 year-1 to 1.54 year-1 to ac-commodate this larger fishery, and balance the model. This P/B value of 1.54 year-1 is real-istic and justified based on recent informa-tion about north Atlantic herring populations (V. Christensen, Fisheries Centre, personal communication, July 1999).  Juvenile Pacific Herring Thomas A. Okey Fisheries Centre, UBC, Canada Robert J. Foy Institute of Marine Sciences University of Alaska, Fairbanks, USA Juvenile Pacific Herring are defined as less than 18 cm in length; age zero to two year olds are considered juveniles. Like adults, juveniles are widely distributed in PWS. Figure 4. Catch history of herring in the PWS area (data provided by J. Wilcock, Alaska Dept. of Fish and Game).  Table 23. Forage fish biomass estimates for PWS, 1995-1997 (data from E. Brown, UAF Institute of Marine Sciences)a. Biomass (t⋅km-2)a,b Year Herring 0 Herring 1 Sandlance Capelin Eulachon 1995 1.693 1.664 0 0.163 0 1996 1.454 9.603 0.196 0.529 0 1997 18.968 0.537 1.590 0.000 3.343 Mean 7.372 3.935 0.595 0.231 1.114 a) Estimates are based on extrapolations from school surface area measurements from airplane surveys and based on empirically-derived assumptions about school packing densities and sub-surface biomass distributions relative to water clarity. Large uncertainties in these factors along with seasonal changes in rela-tive abundance of species compound the uncertainty of these ballpark estimates (E. Brown, UAF Institute of Marine Sceinces, pers. comm.).  b) Estimates of total biomass (t) in PWS can be obtained by multiplying values by 9,059 km2. Table 24. Diet composition (% weight) of herring from 1994-1996 samplesa. Prey categories % diet Herbivorous zooplankton 59.2 Omnivorous zooplankton 32.6 Shallow small epifauna 8.2 a) Functional group composites of the mean pro-portions among years for each taxonomic group; from APEX-SEA data provided by M. Sturdevant (NMFS Alaska Fisheries Science Center). 04080120160198019821984198619881990199219941996Catch (thousands of tonnes)Ye arECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     29 Biomass estimates for herring age 0 (7.372 t⋅km-2) and herring age 1 (3.935 t⋅kg-2) in PWS were developed by E. Brown (unpub-lished data) using areal survey information, with the explicit caveat that these estimates contain considerable uncertainty. The bio-mass estimate for Juvenile pacific herring (13.387 t⋅km-2; including age 2 herring) was then calculated by applying the age 0-1 ratio of biomass (0.54) to age one herring to ac-quire an extrapolated age 2 biomass estimate of 2.125 t⋅kg-2. The biomass estimates for each of the three juvenile age classes were then summed. See Table 23 for the PWS biomass estimates of several forage fish categories. The P/B of 0.54 year-1 and the Q/B of 18 year-1 are taken directly from the adult Pacific herring group. Arrhenius and Hansson (1993) revealed bi-modal distributions for Baltic Sea herring populations (Clupea harengus) at ages zero and four with age zero at almost twice the biomass of age four. The overall ratio of juveniles to adults among eight herring stocks was 0.805, which leads to a juvenile herring estimate of 2.056 t⋅km-2 in PWS when applied to the estimate of adult herring biomass (2.555 t⋅km-2). Interestingly, this estimate is 85% lower than the empirically-based estimate used for PWS (above). The higher (empirically based) estimate is sup-ported by the apparent need for forage fish by the predator biomass in the PWS model.  To calculate diet composition, juvenile her-ring were sampled from the heads of four bays in PWS from October 1995 to Septem-ber 1997 as part of the EVOS Trustee Coun-cil sponsored SEA program. Data missing from unsampled months were interpolated. Fish stomachs were processed at the Insti-tute of Marine Science at the University of Alaska Fairbanks. Prey was identified to the lowest possible taxonomic grouping. Rela-tive proportions of prey groups (nearshore zooplankton) were determined for each fish and pooled for monthly means (Table 25, Appendix 3).  Sandlance Evelyn D. Brown Institute of Marine Sciences University of Alaska, Fairbanks, USA Thomas A. Okey Fisheries Centre, UBC, Canada Sandlance, Ammodytes hexapterus, are schooling zooplanktivores, which burrow into the sand at night. They are found from the intertidal to about 90 m depth and possibly as deep as 275 m. Sandlance are most commonly seen during spring and summer, and may stay buried in the sediment during fall and winter (Love 1996). Estimates of PWS sandlance biomass from 1995 through 1997 have been provided by E. Brown (unpublished data). These estimates are presented in Table 23. A P/B value of 2 year-1 and a Q/B value of 18 year-1 for sandlance are taken from Venier (1996a) for small pelagics in the Strait of Georgia. Sandlance diet composition was summarized from APEX-SEA data provided by M. Sturdevant (Table 27, Appendix 4).  Table 25. Diet composition (% weight) of juvenile herring from 1995-1997a  Taxonomic Group % of diet Herbivorous zooplankton 56.2 Omnivorous zooplankton 41.9 Fish egg 1.9 a) Means of monthly proportions for each taxonomic group; provided by R. Foy and SEA program. Table 26. Diet composition of capelin from 1994-1996 samplesa Prey categories % in diet Herbivorous zooplankton 0.550 Omnivorous zooplankton 0.416 Shallow small epifauna 0.034 a) functional group composites of the mean propor-tions among sampling dates for each taxonomic group; summarized from APEX-SEA data provided by M. Sturdevant  30     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Capelin Evelyn D. Brown Institute of Marine Sciences University of Alaska, Fairbanks, USA Thomas A. Okey Fisheries Centre, UBC, Canada Capelin, Mallotus villosus, is a schooling zooplanktivore that spawns in the intertidal during summer, and which was an histori-cally important food for numerous fishes, birds, and mammals because of its high en-ergy content. However, capelin have be-come much less abundant since the oceano-graphic warming shift of the mid-1970s. Estimates of capelin biomass in PWS for 1995 through 1997 (Table 23) are ballpark estimates, and should be treated as such.  Estimates of 2 year-1 for P/B and 18 year-1 for Q/B are taken from Venier’s (1996a) section on small pelagics in the Strait of Georgia; Table 26 summarizes the diet com-position of capelin. Eulachon Evelyn D. Brown Institute of Marine Sciences University of Alaska, Fairbanks, USA Thomas A. Okey Fisheries Centre, UBC, Canada Eulachon, Thaleichthys pacificus, are pe-lagic schooling smelts that live on the outer continental margin and spawn in fresh water (Love 1996). This species is a reproductive transient in PWS, spending only a brief time as they converge on spawning streams. A mean peak biomass estimate of 1.114 t⋅km-2 was provided by E. Brown for the years 1995-1997 (Table 23). Eulachon did not appear in areal samples in 1995 and 1996, but they ap-peared in large numbers in 1997 (3.343 t⋅km-2). The value entered for biomass, however, is 0.371 t⋅km-2 (1/3 of 1.114 t⋅km-2), to adjust P/B and Q/B to a short PWS residency time.  Values of 2 year-1 for P/B and 18 year-1 for Q/B, used by Venier (1996a) for small pelagics in the Strait of Georgia are also used for eulachon. The diet composition of eula-chon is adapted from the mean of each taxo-nomic group from two APEX-SEA sampling dates provided by M. Sturdevant (NMFS, Alaska Fisheries Science Center, pers. comm.; Table 29, Appendix 4). Fifty percent of the eulachon diet is specified as imported food, assuming that half their food comes from out-side PWS, even considering the above resi-dency time adjustment.  A whole suite of predators feed on eulachon but the extent of this consumption is difficult to estimate due to the ephemeral nature of this species in PWS and the temporal nature of the feeding frenzies that occur when these smelts run. Predators include Pacific cod, sablefish, salmon sharks, spiny dogfish, Pacific halibut, arrowtooth flounder, salmon, baleen whales, orcas, dolphins, pinnipeds, and birds. Indeed, much of the food web partakes directly or in-directly when these summer events occur (E. Brown, UAF Institute of Marine Sciences, pers. comm.; Love 1996). There is some like-lihood that overall consumption of eulachon is underestimated in the model.  Squid Jay Kirsch  Prince William Sound Science Center Cordova, Alaska, USA Thomas A. Okey Fisheries Centre, UBC, Canada Table 27. Diet composition (% weight) of sandlance from 1994-1996 samplesa Prey categories % in diet Herbivorous zooplankton 72.7 Omnivorous zooplankton 21.0 Shallow small epifauna 6.2 a) functional group composites of the mean proportions among sampling dates for each taxonomic group; summarized from APEX-SEA data provided by M. Sturdevant (NMFS, ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     31 An acoustic/trawl survey was conducted from 23-27 January 1997 in the six areas listed below. Echo-square integration of 120kHz sonar signals was used to determine pollock abundance (Table 28). Trawls were used to determine the squid to pollock ratio (estimation of squid to pollock ratios is problematic because the mesh on the net is 4 inches, through which squid may possibly escape, and because the trawls were fished deep (125-215 meters), below the typical nighttime shallow depth distribution of squid. The product of the number of pollock and the squid/pollock ratio is the predicted numerical abundance of squid, which is be-lieved to be an underestimate due to sam-pling bias. Adult PWS squid typically weigh 0.5 kg, although squid were not weighed during this survey. The estimate of 170 t of squid in PWS was converted to 0.019 t⋅km-2 by dividing by 9,059 km2. The P/B and Q/B values used for squid were taken from the Alaska Gyre model (Chris-tensen 1996); these were 3.0 year-1 and 15.0 year-1 respectively. Diet composition infor-mation is adapted from indices of relative importance in prey composition of Loligo opalescens in Karpov and Cailliet (1978). These are shown in Table 30.  LARGER FISHES Walleye Pollock  Mark Willette Alaska Dept. of Fish and Game Cordova, Alaska, USA Walleye pollock (Theragra chalcogramma) spawn in southwest PWS and in Port Bain-bridge during late March. In early June, age-0 pollock (< 15 cm length) typically appear in surface layer net samples in both offshore and nearshore habitats (Willette et al. 1995a, 1996, 1997), where they continue to be found into early September (Willette 1995b). Age 0 pol-lock appear to be ubiquitous in the upper 50 m of PWS waters, and they are also associated with aggregations of moon jellies (Aurelia aurita). Age 1-2 pollock (15 to 30 cm in length) are typically segregated from the adult population during summer when they are commonly encountered in nearshore net sam-ples (Willette et al. 1995a, 1996, 1997). Schools of these age 1-2 pollock migrate through relatively shallow depths (20 – 50m) on the slope outside kelp; they are rarely caught in offshore trawls Adult pollock (age 3+) are greater than 30 cm in length and are commonly captured in sur-face layer (0-50 m) trawl samples in the off-shore areas of the western passages in PWS during May-June (Figure 1) where they feed heavily of the copepod Neocalanus spp. (Wil-Table 28. Squid caught from PWS Pollock surveys. Squid value is thought to be an underestimate for Prince William Sound.  Area No. of Pollock (106)a squid/pollock ratio in trawls Estimated number of squidb South Montague Strait 1.423  0.012 17076  Lower Knight Is Pass 13.369  0.012 160428  Port Bainbridge 7.881  0.091 721  North Montague Strait 3.862  0.012 46344  Green Island 0.304  0.103 31178  Orca Bay 1.492  0.056 83925  Total numbers 28.331  -- 339492     @ ~0.5kg each Biomass                   ~170 t a) Pollock estimates from acoustic data b) assumed to be an underestimate, due to escapement from trawl. Table 29. Diet composition (% weight) of eulachon in PWS, from 1994-1995 samplesa Prey Categories % in diet Omnivorous Zooplankton 99.4 Herbivorous Zooplankton 0.3 Shallow Sm. Epifauna 0.2 a) Functional group composites of the mean proportions among sampling dates for each taxonomic group; summa-rized from APEX-SEA data provided by M. Sturdevant (NMFS, Alaska Fisheries Science Center, pers. comm.). Table 30. Assigned diet composi-tion (% weight) of PWS squids, modified from diet of Loligo opalescens  in Monterey Bay (Karpov and Cailliet 1978). Prey Categories % in diet Off. omni. zoo 96.6Off. herbi. zoo 1.5Deep epifau. 0.4Squid 0.3Near omni. zoo 0.3Eulachon 0.2Shal. sm. epifau. 0.2Shal. sm. infau. 0.2Capelin 0.1Pollock age 0 0.1Near herbi. zoo 0.132     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      lette et al. 1995a, 1996, 1997). After the de-cline of the seasonal bloom of Neocalanus spp., they descend to deeper habitats where they reside for the remainder of the year. During this period, they are distributed throughout PWS, but the biomass is concen-trated in the southwest Sound (NMFS 1993). The majority of the adult biomass during summer is found at depths between 100 and 400m (NMFS 1993).  The total biomass of age 3+ pollock residing in PWS (Table 31) was estimated from the product of the area of the Sound and the mean density of pollock measured during acoustic surveys conducted in western PWS during May-July, 1994 (J. Kirsch, Prince William Sound Science Center, pers. comm.). The minimum biomass of age 3+ pollock was assumed to be that obtained from a bottom trawl survey conducted dur-ing summer in 1989 (NMFS 1993). Bottom trawls likely provide a minimum biomass estimate, because some portion of the pol-lock stock occurs above the bottom where they are not vulnerable to this gear type. The maximum biomass of age 3+ pollock residing in PWS was estimated to be that obtained from acoustic surveys of pre-spawning aggre-gations in southwest PWS (Thomas and Sta-bles 1995, Kirsch 1997). It is unknown what portion of the spawning biomass in southwest PWS resides in the Sound throughout the re-mainder of the year. The biomass of age 0 and age 1-2 pollock was estimated from a simple population model assuming the mean age-specific natural mortality rates used to de-velop the P/B values in Table 31. Annual food consumption (Q/B) was estimated from the annual growth rate of age 0, age 1-2, and age 3+ pollock assuming a gross conversion effi-ciency of 25% (Paul et al. 1988). Annual growth was estimated from mean weight at age of pollock sampled in PWS during May-July, 1994 (Willette et al. 1995a). Mean diet composition of pollock during summer (Table 32) was estimated from samples collected dur-Table 31. Population parameters for walleye pollock in Prince William Sound. Pollock ages Biomass (t⋅km-2) min/ max Catch (t) P/B (year-1) min/ max Q/B (year-1) min/ max Age 0 0.02 0.01-0.05 0 1.28a 0.45-2.34 16.18 12.76-21.97 Age 1-2 0.79 0.39-1.55 0 1.84a 0.90-3.23 3.81 2.01-5.71 Age 3+ 2.20 1.08-4.32 2100b 0.30c -- 2.11 0.41-3.81 Age 1+d 2.99 -- -- 0.707 -- 2.559 -- a) From Bailey et al. (1996); b) From B. Bechtol (Alaska Dept. of Fish and Game, pers. comm.);  c) From Hollowed et al. (1993);  d) Aggregated from the two previous groups; P/B and Q/B values are means weighted by the biomass proportions of those two groups. Table 32. Summer diet composition matrix (in % of volume) for walleye pollock in PWS. Age 1-2 Prey\ Predators Age 0 pelagic demersal composite Age 3+ Age 1+a Herbivorous zooplankton 72.2 37.9 53.3 45.7 0.8 7.6 Omnivorous zooplankton 17.4 10.3 16.8 13.6 38.8 35.0 Carnivorous zooplankton  10.0 -- -- -- -- -- Deep large epibenthic invertebrates -- 1.8 1.4 1.6 33.0 28.2 Shallow large epibenthic inverts. -- 34.2 22.5 28.4 7.0 10.3 Capelin -- 0.2 -- 0.1 12.9 11.0 Juvenile herring -- 0.2 0.4 0.3 0.2 0.2 Pollock age-0 -- 1.7 0.2 0.95 -- 0.1 Pollock age 1-2 -- -- -- -- 2.0 1.7 Juvenile salmon -- 3.6 0.3 2.0 -- 0.3 Squid -- 0.3 1.8 1.1 2.6 2.4 Nearshore pelagic fish -- 9.8 -- 4.9 -- 0.7 Offshore small pelagic fish -- -- 3.3 1.7 2.7 2.5 a) Age 1+ diet composition values are means of the age 1-2 composite and the Age 3+ values, based on the biomass proportion of each group (0.152 and 0.848) ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     33 ing May-July, 1994 (Willette et al. 1995a, 1995b, 1996, 1997). Age 1-2 and age 3+ pollock groups had to be aggregated in the Ecopath model because the Ecosim routine can link only two onto-genetic stages of a species.  These two groups exhibit similar diet composition and probably similar growth rates (relative to the age 0 group). Nevertheless, the age 1-2 fish appear to consume more juvenile fish during summer, whereas the age 3+ fish probably cannibalize age 1-2 fish in winter. Despite these differences, age 0 and age 1-2 groups should not be aggregated, because the age 0 fish are so much smaller and are not preda-tors on other juvenile fish. Perhaps the greatest problem in aggregating the age 1-2 with the age 3+ fish is the cannibalism of the older fish on the younger fish. This may be an important factor in pollock recruitment.    Nearshore Demersal Fishes Tom A. Dean Coastal Resources Associates Vista, California, USA  Nearshore demersal fishes are de-fined as fishes occurring along the shoreline to depths of 20 m and gen-erally found within close association with the bottom. Within Prince Wil-liam Sound nearshore demersals in-clude greenling, sculpins, arctic shanny, gunnels, ronquils, Pacific cod, tomcod, and others (Rosenthal 1983, Laur and Haldorson 1996). Some rock-fish are also considered nearshore demersals, but are treated here as a separate group. Laur and Haldorson (1996) estimated densities of demersal fishes in the Sound in 1990 follow-ing the EVOS (Table 35). Diver surveys were conducted in 4 habitats characterized by dif-ferent vegetation types and exposures: Eel-grass beds in bays, Laminaria and Agarum beds in bays and more exposed points, and Nereocystis beds on very exposed sites in the Sound, especially near the entrances to the Gulf of Alaska. Divers classified fishes in broad size classes (e.g., small and large scul-pins) and there were no estimates of length or length or biomass. We provide rough esti-mates of biomass based on our estima-tion of mean length of fish (T.A. Dean, pers. obs., Table 33) and size-weight relation-ships given in Rosen-thal (1983) and Van Pelt et al. (1997). Species composition varied with habitat and depth, but in all habi-tats the dominant near-shore demersal fishes (by weight) were greenlings and sculpins (Table 33 and Table 34). The mean biomass density within the different habitats ranged from 12 to 46 t ww km-2 (Table 34). On a PWS wide basis, the total biomass of Table 33. Mean biomass per fish (kg wet weight) for nearshore fish in west-ern Prince Sounda.  Taxon Mean weight (kg) Adult cod 0.250 Juv. cod 0.005 Sculpins 0.075 Pholids 0.010 Stichaeids 0.010 Greenling 0.200 Ronquils 0.100 Others 0.020 a Estimates were derived from average lengths of fish within each group, and length-weight relationships given in Rosenthal (1983) and Van Pelt et al. (1997). Counts were based on visual esti-mates from diver observations (T.A. Dean, pers. obs.). Table 34. Biomass (t⋅ww⋅km-2) of nearshore demersal fishes by habi-tat in western Prince William Sound Nearshore demersals Bay Eelgrass Nereocystis Pointsa Adult cod 0.00 5.25 0.00 0.00 Juv. cod 0.47 2.77 0.70 1.28 Sculpins 7.88 0.83 2.03 22.80 Pholids 0.37 0.56 0.01 0.27 Stichaeids 2.03 0.25 0.04 5.26 Greenling 1.00 12.20 16.80 9.00 Ronquils 0.90 0.00 2.00 7.70 Others 0.14 0.06 0.66 0.18 Total 12.79 21.92 22.24 46.49 a. Points are projections of land that define bays Table 35.  Density (No.⋅km-2) of nearshore fishes by habitat, west-ern Prince William Sound. Fish groups and habitats are as defined in Laur and Haldorson (1996). Taxon\Habitat Bays Eelgrass Nereocystis Pointsa Adult cod 0 210 0 0 Juvenile cod 940 5,540 1,390 2,550 Sculpins 1,050 110 270 3,040 Gunnels 370 560 10 270 Arctic Shanny 2,030 250 40 5,260 Greenlings 50 610 840 450 Ronquils 90 0 20 770 Other 70 30 330 090 a. Points are projections of land that define bays 34     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      nearshore demersal fishes was 4.2 t ww km-2. There are no data for P/B or Q/B ratios for nearshore demersal fish in PWS. Based on values for demersal fishes of the Strait of Georgia, Canada, we estimate that the P/B ratio is 1 year-1, and the Q/B ratio is 4.24 year-1. Of the nearshore demersal fishes, only cod are caught in significant numbers by fishers. Since all of these are taken from segments of the cod population that are deeper than 20 m, we consider there to be no nearshore catches. Many smaller fishes are prey to birds (e.g., pigeon guillemots) and larger individuals are prey to river otters (Bowyer et al. 1994). There is also predation by marine mammals and other fish. The diets of the fish vary among species (McConnaughey 1978; Rosenthal 1983, Laur and Haldorson 1995). Most of the diet consists of small invertebrates. There are no quantitative data on diet compositions, but based on general descriptions of diets and relative proportions, we estimate that near-shore demersal fishes consume approxi-mately 70% small epifauna (mostly amphi-pods), 22% large epifauna (mostly crabs), 4% nearshore demersal fish, 2% herbivorous zoo-plankton, 1% shallow small infauna (mostly polychaetes), and 1% sandlance.  Adult Salmon Thomas A. Okey Fisheries Centre, UBC, Canada Adult salmon occur in PWS from June through September as they return from the open sea to their spawning grounds, but a given individual salmon transits the sound in only a few weeks. A residence time of one (1) month was chosen as a reasonable estimate for adult salmon in PWS, considering all species occurring there (L. Huato, UBC Fisheries Centre, pers. comm.). Table 36 shows the av-erage landings of salmon in the sound from 1994 to 1996. The average catch of salmon in PWS for the period 1994-1996 is calculated as 5.373 t⋅km-2 by dividing the estimated annual catch (Table 36) by the total area of PWS (9,059 km2). Table 37 presents minimum estimates of the mean biomass of hatchery and wild pink and chum salmon in PWS from 1994 to 1996, based on wild stock escapement (minimum estimates), hatchery returns, and catches. Run biomass estimates for the three other salmon species in PWS could not be found. However, pink salmon is the dominant spe-cies, contributing about 87% of the catch (in weight), while chum contributes about 10% (from Table 36). Run biomass estimates for the remaining 3% of PWS adult salmon—1359 t of sockeye, 720 t of coho, and 16 t of chinook—were made using a biomass/catch Table 36. Annual landings of salmon in PWS from 1994 to 1996 (commercial and subsis-tence). Species Catcha (N⋅103) Mean weight (kg)b Catch (t) Pink 26319 1.6 42110 Chum 1273 3.9 4965 Sockeye 390 2.8 1093 Coho 145 4.0 579 Chinook 1 11.2 13 Total 28128 (1.8) 48670 a) Mean landings in PWS from 1994 to 1996, (based on Morstad et al. 1997, Appendices E.2); b) Weighted means for 1996 (based on Morstad et al. 1997, Appendix A.5). Table 37. Mean hatchery and wild pink and chum adult salmon runs in PWS, 1994-1996. Stock Estimated pop.  (N⋅103) a Adjusted pop. (N⋅103) b Mean weight (kg) c Biomass of run (t) Biomass/ catch Pink 28987 29571 1.6 47314 1.124 Chum  1658 1735 3.9 6765 1.363 a) Based on Morstad et al. (1997, Appendices E.5 and E.9); b) Adjusted to account for 30% of wild stock escapement into non-index streams (B. Bue, Alaska Dept. of Fish and Game, pers. comm.); c) Weighted means in 1996 (based on Morstad et al. 1997, Appendix A.5). ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     35 ratio of 1.24 derived from pink and chum ratios shown in Table 36 and Table 37. Summing the estimates for each species gives a mean estimate for PWS salmon runs (wild and hatchery) including catches from 1994 to 1996 of 56,174 t, or 6.201 t⋅km-2⋅year-1. The annual biomass estimate of 0.517 t⋅km-2 was then calculated by dividing this peak estimate by 12, assuming that a given individual salmon has a PWS resi-dence time of one month.  However, this method underestimates salmon in PWS because of observer ineffi-ciency when developing escapement indices. Moreover, run biomass estimates are only a portion of total population biomass in PWS because they do not account for predation on salmon while transiting PWS toward spawn-ing streams. For these reasons, the 1994-1996 run estimate was doubled to arrive at an adult peak salmon biomass estimate of 12.402 t⋅km-2 and a corresponding annual biomass estimate of 1.034 t⋅km-2, used as the entered value. This corrected, one twelfth (1/12) adult salmon biomass estimate was entered in the biomass category while the remaining 11/12 was entered in the immi-gration column, and 75% of that value was entered as export (the alternative is that the difference is entered as net import). The adult salmon group poses a difficult problem for the Ecopath model since they are summer transients and feed little while in PWS, if at all. Moreover, their transient nature inhibits useful and accurate calcula-tion of P/B and Q/B values for their within-PWS adult stage. However, instantaneous mortality rates across all life stages have been calculated by Bradford (1995), and his PWS-weighted mean of 6.476 year-1 can be used for an adult salmon P/B (see Table 38), even though it is unlikely that P/B is evenly distributed throughout the life cycle.    A Q/B value of 12 year-1 applies to pink salmon in the Alaska gyre (L. Huato, UBC Fisheries Centre, pers. comm.; Table 10 in Christensen 1996), and is used for the adult salmon group. The annual Q/B and P/B val-ues may not relate to the role of adult salmon during their spawning stage, but this does not affect the other components of the model (excepting a few that eat adult salmon) because 100% of the food of PWS adult salmon is assigned as import representing the imported secondary production from the Alaska gyre in the form of salmon growth during that ontogenetic stage.   Adult salmon are mostly eaten by resident orcas and eagles (bears are not in the model since most bear feeding occurs in rivers, out-side the PWS system as defined here). Salmon that successfully spawn and die before being consumed are considered to become 'nekton falls' (one of the three detritus categories) upon dying, but only a portion of the carcasses of the successful spawners make it back to the PWS system to become 'nekton falls' (the PWS system extends to the upper intertidal, but not up rivers). Most of the unused portions of the 'nekton falls' detritus become inshore detritus, the unused portion of which becomes offshore detritus, which can ultimately be ex-ported from the system.   Adult salmon probably eat very little in PWS as they return to spawning grounds. This is reflected in the Ecopath model by giving them 99% ‘imported’ food (in the diet composi-tion). The remaining 1% of their food is as-signed to herring, sandlance, eulachon, and capelin to achieve a realistic trophic ranking for salmon. Table 38. Total mortality (Z) for five species of salmon in PWSa. Species Proportions of biomass caught Total mor-tality (Z; year-1) Pink 0.865 6.33 Chum 0.102 7.59 Sockeye 0.022 6.55 Coho 0.012 6.40 Chinook 0.000 6.76 Weighted meanb -- 6.48 a) From Bradford (1995); b) Weighted by the catch proportions. 36     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Rockfishes Thomas A. Okey Fisheries Centre, UBC, Canada Rockfishes, family Scorpaenidae, include over 100 species worldwide including 64 from the northeast Pacific Ocean (Orr et al. 1992). The vast majority of Rockfishes are in the genus Sebastes, and are demersal groundfish (often in rocky bottom habitats), but some are classified as pelagic, as they are found in mid-water or near kelp cano-pies. Rockfishes occupy a variety of niches, but they are presented as an aggregated group in this model. The total annual rockfish biomass estimate of 0.254 t⋅km-2 in PWS was derived by mul-tiplying the mean PWS landings from 1994 to 1996 by the mean biomass/commercial landings ratio from the greater Gulf of Alaska management region for that time pe-riod (Table 39). The accuracy of this bio-mass estimate is questionable because rec-reational landings of pelagic and demersal rockfishes in PWS often exceeded commer-cial landings during the period in question (See below).  Estimates of rockfish P/B (0.17 year-1) and Q/B (3.44 year-1) were taken from Dalsgaard et al. (1998). A generalized diet composition for rockfish was derived by adapting information in Yang (1993) for six spe-cies of commercially im-portant rockfish that occur in PWS and from Rosen-thal et al. (1980) for two other PWS rockfish spe-cies. The diet of a ‘com-posite’ rockfish was de-rived by taking the aver-age of each prey category and using these values as diet proportions (Table 40). Weighted means among species were not used because ratios of es-timated exploitable bio-masses among species in the Gulf of Alaska were not expected to relate directly to the rela-tive densities of these species in PWS. Fur-thermore, the resulting diet composition for rockfish is skewed towards those of the com-mercially important slope and shelf rockfishes and away from other rockfishes such as those that occur in more shallow habitats, or recrea-tionally important species.  A total landings estimate of total rockfishes in PWS from 1994-1996 (89.255 t, or 0.010 t⋅km-2) is the sum of commercial landings (Table 39, Table 43) and recreational landings (Table 75).  Parameterization of PWS Rockfish (Sebastes) indicates that this composite group is declin-ing in PWS, though this indication is based on limited information. A immigration term of (-0.14 t⋅km-2⋅year-1) was used to artificially bal-ance this group by adding immigrating adults to the population every year. This term was added to the migration of rockfish in contrast to our default assumption that the 'composit' rockfish in PWS has no net migration. This 'trick' was used to balance the rockfish group because there was no justification for increas-ing the given biomass and P/B values, though they contained uncertainty, and because a 'balanced' model, without explicit declines, is convenient for exploring trophic relationships. Table 39. Biomass estimates for PWS rockfish for 1994-1996. Year Biomass /landingsa (t/(t⋅year-1) PWS land-ings (t⋅year-1)b Estimated PWS Bio-mass (t) Estimated PWS Biomass (t⋅km-2) Slope means 31.4 30.571  960.9  0.106 1994 36.6 25.897  947.8   1995 27.2 42.881  1166.4   1996 30.5 22.207  677.3   Pelagic means 23.2 9.027  209.3  0.023 1994 25.6 7.995  204.7   1995 19.4 12.707  246.5   1996 24.5 6.165  151.0   Demersal means 78.9 14.375  1134.6  0.125 1994 77.9 10.032  781.5   1995 61.3 13.512  828.3   1996 97.6 19.240  1877.8   Grand Totals 44.5 53.973  2304.8  0.254 a)  Biomass/landings ratios pertain to exploitable biomass and commercial landings in the Gulf of Alaska management region (from NPFMC 1995 and 1997); b)  Commercial landings information from B. Bechtol (unpublished data). ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     37 The other alternative is to add a negative biomass accumulation term to explicitly dis-play the indicated declines in biomass for this composite group. The upshot of this balancing 'trick' is that the negative biomass accumulation value (-0.14 t⋅km-2⋅year-1) represents rockfish decline at that rate.   Nearshore Rockfish  Tom A. Dean Coastal Resources Associates Vista, California, USA Diver surveys in the nearshore (less than 20 m) in 1990, 1996, and 1997 found few rockfish in this zone (T. Dean and S. Jew-ett, unpublished data). Only juvenile copper rockfish were at all abundant. Densities (No. per 100 sq. m) in various habitats are shown in Table 42. Our estimates are probably low for Nereo-cystis since we sampled only at relatively shallow depths there (less than 10 m) and most rockfish are deeper. Our data suggest that rockfish make up a very small propor-tion of the biomass of nearshore benthic fishes. Rosenthal (1980) indicated that rockfish can be very abundant. However, most of his data were from Nereocystis beds on the outer margins of the Sound (Danger Island, Schooner Rocks, Zaikof Point). These are sites on the margin, or in some cases ex-cluded from our current PWS boundary defi-nition. These are very special habitats and not very well represented within the Sound proper. (We estimate that Nereocystis habitat makes up less than 1 % of total nearshore habitat). While Rosenthal found rockfish at depths less than 20 m in these habitats, they Table 40. Diet compositions (in % weight) of rockfish species in PWSa Prey Group POP Rougheye Northern Dusky s-spine TH Shortraker China Black Mean Omnivorous Zooplankton 89.7 6.0 96.6 72.3 0.7 -- -- 56.7 40.3 Deep epifauna 6.0 65.7 0.2 7.8 80.3 -- 70.0 0.2 28.8 Squid 8 20.0 0.3 6.2 0.7 82.0 -- -- 13.8 Shallow small epifauna -- -- -- -- -- -- 15.0 7.2 2.8 Deep demersals 1.9 6.9 0.4 -- 15.5 -- -- -- 3.1 Myctophids -- -- -- -- -- 18.0 -- -- 2.3 Herbivorous Zooplankton 1.7 -- 2.5 13.7 -- -- -- 1.0 2.4 Sandlance -- -- -- -- -- -- -- 32.8 4.1 Shallow large epifauna -- -- -- -- -- -- 15.0 0.2 1.9 Deep Infauna -- 1.4 -- -- 1.2 -- -- -- 0.3 Nearshore demersals -- -- -- -- -- -- -- 1.4 0.2 Age 0 pollock -- -- -- -- 0.9 -- -- -- 0.1 Capelin -- -- -- -- 0.7 -- -- -- 0.1 Herring -- -- -- -- -- -- -- 0.5 0.1 a) Diet compositions for China and Black rockfishes were adapted from prey indices in Rosenthal et al. (1988). Diet compositions for all other species are from Yang (1993).  Table 41. Estimated biomass of demersal fish species in PWS, 1989a. Species Biomass (t) Fishes    Walleye Pollock 7140   Pacific Cod 2040   Sablefish 1470   Arrowtooth Flounder 19300   Flathead sole 3000   Rex sole 1510   Skates 3402   Halibut 1880   Rougheye Rockfish 844 Invertebrates    Shrimp 101   Tanner Crab 2200 a)  NMFS (1993) Table 42. Densities of juvenile cop-per rockfish in different PWS habi-tats. Habitat No.⋅km-2 Eelgrass beds  0.002 Bays  0.004 Points  0.079 Nereocystis  0.027 38     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      were more abundant in deeper waters. According to Love (1996), copper rockfish eat plankton, and adults eat octopods, shrimps, crabs, and small fishes. This is similar to the diets of some of the commer-cially-important species used to generate the diet composition of a generalized rockfish (see previous section).  Miscellaneous Demersal Fishes Thomas A. Okey Fisheries Centre, UBC, Canada Biomass information was limited for some species of explicitly defined fishes. Two methods were used to derive preliminary estimates of biomass: (1) the mean PWS landings for a particular species from 1994 to 1996 was multiplied by the mean bio-mass/landings ratio from the greater Gulf of Alaska management region for that period, (2) biomass estimates from a 1989 multi-species trawl survey (post spill) were ex-tracted from appropriate areas and used as a proxy for 1994-1996 estimates or for com-parison to the results of method (1). These estimates are shown in Table 41 and Table 43, and in the following sub-sections. P/B and Q/B values taken from other models are shown in Table 44.  Deep Demersals (skates and flatfishes)  For the purposes of this deep-demersal group, flatfishes include flathead sole and rex sole, and skates include big skate, Aleu-tian skate, and Alaska skate. The biomass es-timate of 0.873 t⋅km-2 is based on the sum of post-spill 1989 biomass estimates of these flatfishes and skates listed in Table 41 above (from NMFS 1993). Biomass/landings ratios from the Gulf of Alaska region were not used to convert landings to biomass in this case because catches of this group were increasing during this period. For example, the commer-cial fishery landings in PWS for flounders and skates increased from zero in 1994 to 18.3 t, or 0.002 t⋅km-2, in 1996.  Estimates of P/B (0.775 year-1) and Q/B (3.21 year-1) were calculated by taking the mean of the flatfish estimates in Dalsgaard et al. (1998). Diet composition values were esti-mated from considerations in Love (1996).  Pacific Cod Pacific cod (Gadus macrocephalus) is a schooling species found near soft or gravel bottoms mostly between 45 m and 275 m. They spawn in deeper water, but move shal-low to feed during late spring and summer, particularly the juveniles (Love 1996).  The Pacific cod biomass esti-mate of 0.555 t⋅km-2 in PWS was derived by multiplying the mean PWS landings from 1994 to 1996 by the mean exploitable biomass/landings ratio of Pacific cod from the greater Gulf of Alaska management region for that time period (Table 45). This Table 43. Reported PWS commercial landings (unpublished data provided by B. Bechtol, Alaska Dept. of Fish and Game).              Landings (t) Group 1994 1995 1996 Mean Lingcod 4.662 1.298 2.962 2.998 Pacific Cod 752.184 708.362 307.863 594.185 Sablefish 126.249 254.295 116.045 166.854 Pollock 2.570 2947.865 1659.676 1548.997 Flounders 0 1.584 11.156 4.281 Skates 0 1.072 7.120 2.753 Sharks (all) 0 0.158 9.014 3.081 Rockfish 43.924 69.100 47.612 53.974    Pelagic 7.995 12.707 6.165 9.027    Demersal 10.032 13.512 19.240 14.375    Slope 25.897 42.881 22.207 30.571 Table 44. P/B and Q/B values from other Ecopath models Group P/B (year-1) Q/B (year-1) Pacific Cod 1.200a 4.000 a Sablefish 0.566 a 6.420 a Lingcod 0.580b 3.300 b Other Flounders 0.775 b 3.210 b All Rockfish 0.170 b 3.440 b a) from Livingston (1996), for the Bering Sea ecosystem; b) Dalsgaard et al. (1998), for the Strait of Georgia ecosys-tem. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     39 estimate is 247% greater than the 0.225 t⋅km-2 estimate adapted from the 1989 post-spill PWS multi-species trawl survey data (NMFS 1993). Nevertheless, 0.555 t⋅km-2 could well be an underestimate since the estimation method does not account for pre-recruits.  A P/B value of (1.2 year-1) and a Q/B value of (4.0 year-1) were taken from Livingston (1996) estimates for the southern BC shelf model. Diet composition information in Table 46 was adapted from Yang (1993). Shallow and deep prey allocations were made by an assumed 25% feeding in areas shallower than 20 m. Note that 12.5% of the prey of Pacific cod is fishery discards. This prey item was added to the Pacific cod diet as an import (see Sablefish discussion).   Lingcod Lingcod, Ophiodon elongatus, are large predatory greenlings (Hexagrammidae) that live mostly on or near the bottom in rela-tively shallow water feeding on fishes, squids, and octopods, or guarding large egg masses if male. Lingcod, like rockfish dis-cussed above, have both commercial and recreational importance in PWS. Explicit identification of these groups in the model may help provide insights into the impacts of changing exploitation levels as user demographics change in the future. The mean recreational landings of lingcod in PWS between 1994 and 1996 was 20.6 t⋅year-1 Adding this to the mean commercial landings of 3 t during the same period gives an annual fishery extracted biomass of 23.6 t (0.003 t⋅km-2). Application of a conservative biomass/catch ratio, like that for Pacific cod (8.5), results in prelimi-nary lingcod biomass estimate of 200.6 t in PWS, or 0.022 t⋅km-2. The resulting lingcod estimate may not be re-alistic as the actual ling-cod biomass/landings ratio is likely different than that of Pacific cod. A PWS biomass of 0.022 t⋅km-2 is probably an underestimate for lingcod. Estimates of lingcod P/B (0.58 year-1) and Q/B (3.3 year-1) were taken from Dalsgaard et al. (1998). Diet composition values for ling-cod are adapted from a discussion in Cass et al. (1990).   Sablefish Sablefish (Anoplopoma fimbria) are schooling fish that live on or near muddy or sandy bot-toms from 180 m to over 900 m when adult (Love 1996). Juveniles are found shallower than 180 m. The sablefish biomass estimate of 0.195 t⋅km-2 in PWS was derived by multiplying the mean PWS landings from 1994 to 1996 by the mean biomass/landings ratio from the greater Gulf of Alaska management region for that period (Table 47). This estimate is 20% greater than the 0.162 t⋅km-2 adapted from the 1989 post-spill PWS multi-species trawl survey data Table 45. Biomass estimates for PWS Pacific cod for 1994-1996. Year Biomass/ landingsa PWS landingsb Estimated PWS biomass (t) Estimated PWS biomass (t⋅km-2) 1994 6.231 752.184 4686.859 0.517 1995 8.417 708.362 5962.283 0.658 1996 10.726 307.863 3302.139 0.365 Mean 8.458 594.185 5025.617 0.555 a) Biomass/landings ratios apply to the Gulf of Alaska management region and are derived from NPFMC (1995 and 1997); b) Landings information from B. Bechtol (unpublished data). Table 46. Diet composition (i% weight) of Pacific Coda Prey group % in diet Discards 12.5 Shallow epifauna 11.9 Deep epifauna 35.6 Infauna 5.3 Squid 2.5 Shallow demersal 7.9 Deep demersal 7.9 Capelin  1.9 Arrowtooth 5.8 Pollock 7.4 Eulachon 0.3 Herring  0.4 a) From Yang (1993). 40     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      (Table 41), but it is likely still an underesti-mate since exploitable biomass is only a fraction of total biomass.  The P/B value used for Sablefish in PWS (0.566) was derived by taking the mean of juvenile and adult mortality estimates weighted by the proportions of juvenile and adult biomasses used for the southern BC shelf model by Livingston (1996). McFarlane and Beamish (1983) presented a natural mortality rate for juveniles between age 0 and age 4 of 0.6 year-1, and Stoker (1994) presented a natural mortality rate for adults of 0.08 year-1. The proportion of juvenile to adult biomass on the South-ern BC shelf is 15:1. Like-wise, a Q/B value of 6.42 year-1 was calculated by tak-ing the weighted mean of the Q/B values given for juve-nile and adult sablefish by Livingston (1996). There-fore, derivation of sablefish P/B and Q/B values for PWS is based on the assumption that proportion of juvenile to adult biomass is the same there as on the southern BC shelf. The mean commercial landings for the three years was 0.018 t⋅km-2. Diet composition for sablefish in Table 48 was based on Yang’s (1993) study in the Gulf of Alaska. This study also revealed that sable-fish consumed all size classes of pol-lock (0, 1-2, and 3+). Equal alloca-tion among pollock age classes is based on the assumption of equal proportions consumed by sablefish. Note that 29% of sablefish prey is fishery discards based on information from the Gulf of Alaska. The 'nekton falls' group, which include discarded fish carcasses, are fed to Sablefish, but some food is imported into their diet, reflecting consumption outside the system.   Arrowtooth Flounder Mark Willette Alaska Dept. of Fish and Game Cordova, Alaska, USA Arrowtooth flounder (Atheres-thes stomias) is a large preda-tory flatfish which may be the single most abundant fish spe-cies in the Gulf of Alaska (Wil-derbuer and Brown 1993). The biology of arrowtooth flounder is not well known. Sexual ma-turity occurs at a length of about 30 cm and spawning off the coast of Washington takes place during winter (Rickey 1995). Larvae occur in the sur-face layer 0-200 m during summer (Taylor 1967) where they feed on copepods and eggs (Barraclough and Fulton 1968). The biomass of arrowtooth in PWS appears to have increased substantially from 4,000 to 40,000 tons be-tween 1978 and 1989 (Parks and Zenger 1979, NMFS 1993). In 1989, arrowtooth comprised about 55% of the bottomfish biomass in the Table 47.  Biomass estimates for PWS sablefish 1994-1996. Year Biomass/ landingsa PWS landingsb PWS  Biomass (t) Biomass (t⋅km-2) 1994 10.216 126.249 1289.760 0.142 1995 10.885 254.295 2768.001 0.306 1996 10.610 116.045 1231.238 0.136 Mean 10.570 166.854 1763.647 0.195 a. Biomass/landings ratios apply to the Gulf of Alaska management re-gion and are derived from NPFMC (1995 and 1997); b. Landings information from B. Bechtol (unpublished data). Table 48. Diet composition (% weight) of sablefisha Prey group % in diet Discards 29.1 Age 0 Pollock 10.5 Age 1-2 Pollock 10.5 Age 3 Pollock 10.5 Squid 8.0 Omni. Zoopl. 6.7 Eulachon 5.5 Jellies 5.4 Deep epifauna 5.1 Deep demersal 4.8 Herring 2.2 Pacific Cod 0.8 Infauna 0.4 Capelin 0.3 a. Compiled from Yang (1993). Table 49.  Population parameters for arrowtooth flounder in PWS.  Arrowtooth stages Biomass (t⋅km-2)a Biomass min-max a P/B (Z; year-1)b Q/B (year-1)c Juveniles 0.57 0.08-1.05 0.22 3.03 Adults 4.00 0.60-7.36 0.22 3.03 a. NMFS (1993); b. Wilderbuer and Brown (1993); c. Smith et al. (1991). ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     41 Sound (NMFS 1993). Ecopath input pa-rameters for arrowtooth flounder are pre-sented in Table 49. Despite its high abun-dance, there is no directed fishery on arrow-tooth flounder in the PWS region.  Approximately 56% of the juvenile biomass and 80% of the adult biomass of arrowtooth flounder occurs in southwestern PWS (see Appendix 6). The remainder of the juvenile biomass is found in Orca Bay and Port Fi-dalgo (NMFS 1993). Juveniles tend to be distributed between 20 and 200m; whereas, adults typically occur between 100 and 400m. In the present study, minimum and maximum biomass densities were calculated from the lower and upper 95% confidence intervals on the mean biomass estimated during a 1989 trawl survey (NMFS 1993). The annual food consumption of arrowtooth was estimated from a laboratory analysis of the energetics of yellowfin sole (Smith et al. 1991). Size at age data indicate that arrow-tooth and yellowfin sole exhibit very similar growth rates in the Gulf of Alaska (Wilder-buer and Brown 1993). The diet of juvenile arrowtooth during summer is dominated by shrimp and capelin, whereas adult diets are dominated by walleye pollock (Table 50).   Pacific Halibut International Pacific Halibut Commission staff Seattle, Washington, USA  We estimated biomass (total biomass of age 8 and older, round weight) by calculating rela-tive abundance from catch/effort and bottom area for Prince William Sound and IPHC Area 3A, and scaling to the 3A biomass (see Table 51). Density (catch/effort) times area is usu-ally proportional to abundance, and the ratio of relative abundance for the two areas times absolute abundance in one area (3A) equals absolute abundance in the other area (PWS). This calculation must be treated with caution, because the catch/effort data in PWS are very limited. The total catch in PWS is only 1.5-2.5% of the 3A catch, and the number of ship logs collected for catch/effort in PWS is a low proportion of the total landings there. The ra-tio of PWS biomass to 3A biomass is similar to the ratio of PWS catch to 3A catch, so we think the estimates are in the right ballpark. However, we would be inclined to use the es-timated values as a range, rather than a trend. The mean biomass of Pacific halibut in PWS for the period 1994-1996 is 6133 t, or 0.677 t⋅km-2 (Table 51). Our estimate of total mortality (Z) comes from adding the calculated fishing mortality (F) to a constant estimate of M (=0.2 year-1); i.e., 1993: 0.33; 1994: 0.34; 1995: 0.30; 1996: 0.32; 1997: 0.34. The mean Z for the 1994-1996 period, 0.32 year-1, is used as the P/B estimate. A Q/B of 1.73 year-1 was taken from Venier (1996b) who derived it from an em-pirical equation in Christensen and Pauly (1992, p. 14). This value is close to the mean of the following two Q/B values suggested by P. Livingston for the gulf of Alaska (unpub-Table 50.  Summer diet composition matrix (% weight) for arrowtooth flounder in PWS (from Yang  1993). Prey\Predator Juveniles Adults Omnivorous zooplankton 15.0 3.0 Deep large epifauna 25.0 4.0 Capelin 44.0 14.0 Juvenile herring 5.0 7.0 Adult herring 5.1 4.0 Pollock age-0 3.7 21.8 Pollock age 1-2 1.2 45.2 Squid 1.0 1.0 Table 51.  Biomass of Pacific halibut in PWS  Year PWS catch (t) 3A catch (t) No. skates PWS catch/effort (t/skate) 3A catch/effort (t/skate) 3A  biomass (t)PWS catch fraction PWS  biomass (t) PWS  biomass fraction Density (t⋅km-2) 1993 214 13753 259 0.1071 0.2364 320508 0.0156 7476 0.0233 0.825 1994 220 15023 213 0.1084 0.1997 291199 0.0147 8140 0.0280 0.899 1995 214 11092 435 0.0780 0.2357 275251 0.0193 4691 0.0170 0.518 1996 295 11909 910 0.1110 0.2675 260614 0.0248 5569 0.0214 0.615 1997 366 14926 1532 0.0824 0.2639 238751 0.0245 3839 0.0161 0.424 a. Area 3A extends along the continental shelf from Cape Spencer to the west end of Kodiak Island, including PWS. 42     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      lished data): Halibut consume 0.4% body weight daily for individuals 50-79 cm in length (0.004⋅365 = 1.46 annual ration Q/B for Ecopath) and 0.3% body weight daily for individuals greater than 80 cm in length (0.003⋅365 = 1.095 annual ration Q/B for Ecopath). The diet composition of Halibut in Table 52 is adapted from Yang (1993). Halibut occur from depths of 6 m to over 1,000 m, and generally migrate to shallow waters during summer to feed (Love 1996). The deep/shallow zone allocations were based on the assumption that 25% of demersal fish and epifaunal biomass was taken in waters shallower than 20 m. Yang (1993) showed that 57.4% of GoA halibut diet in the sum-mer of 1990 was walleye pollock, and that all three size classes were consumed. Biomass of consumed pollock was thus allocated equally among pollock age classes in the model. However, only ju-venile arrowtooth flounder (<30 cm) were consumed by halibut.    The commercial catches of halibut in PWS are given in Table 51, and the rec-reational catches are given in Table 75. The total catch estimate is the grand mean of the sum of recreational and commencial catches from 1994-1996: 626 t, or  0.069 t⋅km-2.  Sharks Lee Hulbert NMFS Auke Bay Laboratory Juneau, Alaska, USA  This group is composed of salmon sharks (Lamna ditropis), spiny dogfish (Squalus acanthias), and sleeper sharks (Somniosus pacificus).  Personal observations and anecdo-tal evidence suggest that shark abundance has increased dramatically throughout the 1990s. Anecdotal accounts of increasing numbers of dogfish in PWS are supported by a time series of relative abundance (CPUE) for dogfish cal-culated from International Pacific Halibut Commission longline survey data (Figure 5; data provided by IPHC biologist Dan Ran-dolf).  Currently there are no quantitative estimates of biomass for these species in PWS.  Given the potential trophic and ecological impor-tance of these predators in PWS, research is needed to obtain more realistic estimates of  biomass, abundance, and diet composition. The estimates in this section were made to provide input parameters for the Ecopath modelling exercise, but caution is advised when considering the usefulness of these pre-liminary estimates for other purposes, as some are little more than rough approximations.   Figure 5. Average annual shark bycatch per 100 hooks.  Compiled from IPHC longline survey data collected in the GoA between Nuka Point and Cape St. Elias. Table 52.  Diet Composition (% weight) of Halibuta Prey group % in diet Pollock 57.4 Deep epifauna 14.1 Discards 7.1 Salmon 5.3 Shallow epifauna 4.7 Juv. Arrowtooth 4.2 Deep demersal 3.8 Shallow demersal 1.3 Pacific Cod 1.0 Capelin  0.9 Squid 0.2 a) Adapted from Yang (1993).  05101520251984 1985 1986 break 1993 1994 1995 1996 1997 1998YearMean CPUE(fish / 100 hooks)Spiny dogfishSalmon sharkSleeper sharkUnidentified sharksECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     43 Biomass estimates of four thousand tonnes of  salmon sharks in Prince William Sound (0.442 t.km-2), and one thousand tonnes each of spiny dogfish and sleeper sharks (0.11 t.km-2) resulted in an overall shark biomass estimate of 0.662 t.km-2.  Derivations for Q/B were based on daily ration estimates of 1% body weight per day for spiny dogfish and sleeper sharks, and 5% body weight per day for salmon sharks.  These Q/B estimates were corrected using peculative estimates of annual residency durations of 180 days for salmon sharks and spiny dogfish, and 365 days for sleeper sharks. A weighted average of these adjusted Q/B estimates were then calculated based on the relative biomass es-timates for each species. The derivation of Shark Q/B estimates are shown in Table 54. The P/B estimate of 0.1 year--1 for these spe-cies are based on a natural mortality esti-mate for spiny dogfish (Polovina 1996). Immigration is assumed to ap-proximately equal emigration.  Commercial shark bycatch in Alaska waters is poorly docu-mented.  Shark bycatch is fre-quently recorded as ‘unidentified shark’, ‘shark’, or ‘other fish.’ To account for fishery removals of PWS sharks, a discard flow of 0.0038 t⋅km-2⋅year-1 was equally apportioned between commercial and recreational fisheries. This flow represents 10% of the pro-duction of PWS sharks (the prod-uct of the specified biomass and P/B estimates), corrected with the overall shark residency multiplier (0.58; Table 54). Dividing the resulting flow equally among commercial and recreational fish-eries results in an estimate (0.0019 t⋅km-2⋅year-1 each) that is very close to the speci-fied estimate for shark discards associated with commercial fisheries in PWS (0.0022 t⋅km-2⋅year-1; Table 76). Thus, the former value was used for recreational discards, and the latter for commercial, resulting in an adjusted total specified shark discard of 0.004 t⋅km-2⋅year-1. Incorporation of these catch estimates are needed to avoid incon-sistency in model parameterization (i.e., not including parameters of the model we know are not zero).  In 1997 the Alaska Board of Fisheries closed all commercial shark fishing and heavily regu-lated the sport fishery in Alaska state waters.  No Federal Management plan exists specifi-cally for sharks in the Gulf of Alaska and the Aleutians. Sixgill and blue sharks also occur in the PWS area but are not explicitly in-cluded in this exercise, i.e., their biomass is assumed to be part of the overall 'shark' bio-mass estimates specified herein. The salmon shark, a large pelagic predator, is the sister species to the better known porbea-gle (Lamna nasus), and is also closely related to the white shark (Carcharodon carcharias), and mako sharks (Isurus oxyrinchus and I. paucus).  Salmon sharks live at least 25 years Table 53. Prey items and % diet composition of sleeper shark (Somniosus pacificus), collected in Gulf of Alaska in 1996 (Mei-Sun Yang,  NMFS, Alaska Fisheries Science Center, pers. comm.). Sleeper shark prey Frequency (%) % in diet Gastropoda (snail) 9.09 0.49 Fusitriton sp. (snail) 9.09 0.19 Cephalopoda (squid and octopus) 27.27 0.17 Teuthoidea (squid) 36.36 0.62 Octopus dofleini (octopus) 72.73 4.63 Crangonidae (shrimp) 9.09 0.01 Pagurid (hermit crab) 9.09 0.01 Teleostei (unidentified fish) 45.45 0.33 Oncorhynchus sp. (salmon) 9.09 4.49 Gadidae (gadid fish) 9.09 0.49 Theragra chalcogramma (walleye pollock) 9.09 5.22 Atheresthes stomias (arrowtooth flounder) 63.64 67.2 Sebastes sp. (rockfish) 9.09 2.06 Pleuronectid (unknown flatfish) 18.18 0.86 Hippoglossoides elassodon (flathead sole) 9.09 0.98 Fishery offal 9.09 12.3 Table 54. Biomass and Q/B estimates for PWS Sharks. Population parameters Salmon sharks Sleeper sharks Spiny dogfish PWS sharks Biomass (t) 4000 1000 1000 -- Daily ration (%) 5 1 1 0.58 Residency time (year) 0.5 1 0.5 -- Overall Q/B (year-1) 18.25 3.65 3.65 7.00 PWS-adjusted Q/B (year-1) 9.13 3.65 1.82544     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      and average size range in PWS appears to be between 180-230cm total length (Lee Hul-bert, unpublished data).  Length and age at maturity estimated to occur at 140cm PCL and 5 years for males, and 170-180cm PCL and 8-10 years for females. L. ditropis is ovoviviparous with an annual fecundity of up to 5 pups (Tanaka 1980); the gestation period is not documented.  Based on mating occurring in the late summer and parturition occurring in the spring, gestation may be around 9 months (K. Goldman, Virginia In-stitute of Marine Science, pers. comm.).  Salmon sharks are opportunistic predators.  Their diet includes salmonids (Oncorhyn-chus), rockfish (Sebastes), lancetfish (Ale-pisaurus), daggertooth (Anotopterus), sable-fish (Anoplopoma), spiny dogfish (Squalus acanthias), lumpfishes (Cyclopteridae), lanternfishs (Myctophidae), sculpins (Cotti-dae), pollock (Theragra chalcogramma), Pacific tomcod (Microgadus proximus), her-ring (Clupeidae), halibut (Pleuronectidae), squid (Teuthoidea), and benthic crustaceans (Nagasawa  1998, Tanaka 1980, 1986, Cas-tro 1983, Sano 1960, 1962). Temporal abundance of salmon sharks in PWS is not documented.  Distribution and abundance of salmon sharks are associated with aggrega-tions of prey (Blagoderov 1994) and have been observed in spring (April-May) during the sac roe herring fishery and fall (Septem-ber-October) during the herring bait fishery.  Peak abundance appears to occur during July and August, corresponding with the return of adult salmon to PWS.  Neave and Hanavan (1960) observed no obvious pat-tern of change in distribution of salmon sharks in the Gulf of Alaska between May and Sep-tember. An occasional salmon shark is taken in trawl gear during the PWS winter pollock fishery (Robert Bercelli, Alaska Dept. of Fish and Game, pers. comm.).  Table 55 lists the frequency of occurrence of salmon shark prey taxa in 11 stomachs col-lected from mid to late July, 1998 (K. Gold-man, Virginia Institute of Marine Science, unpublished data).  Eight sharks were col-lected in Montague strait, two in Aialik Bay, and one was collected in Resurrection Bay.  Weight data for individual prey was unavail-able; breakdown of diet composition by per-cent biomass is based on estimated weights of prey taxa. Spiny dogfish are adaptable predators that often congregate in packs.  They can grow to 130 cm and over 9 kg.  Dogfish age and length at maturity vary greatly with region, and have been estimated to range from 16-35 years and up to 94 cm for females (Love 1996, Smith 1998). They are ovoviviparous and average 7 pups per parturition.  Gestation period is the longest of any vertebrate at 22-24 months (Saunders and McFarlane 1993).  Diet composition of dogfish in PWS has not been documented, but diet composition information is available from British Columbia during the 1970s (Table 56). They are known to prey heavily on schools of spawning capelin, and aggregations of dogfish are often associated with herring returning to coastal waters of British Columbia.  Principal food appears to be herring (Clupeidae), sandlance (Ammo-Table 55. Derivation of estimated diet composition of salmon sharks in PWSa  Salmon shark prey taxa Frequency in 11 stomachs Frequency (%) Mean weight (kg) Biomass  (%) Salmonids (Oncorhynchus) 5 26 2.2 40 Sablefish (Anoplopoma) 5 26 2 36 Pollock, Cod (Gadidae) 1 5 1 4 Rockfish (Sebastes) 1 5 0.3 1 Herring (Clupeidae) 2 11 0.05 0.4 Spiny dogfish (Squalus acanthias)  1 5 2 7 Squid (Teuthoidea) 3 16 0.1 1 Halibut (Pleuronectidae) 1 5 3 11 a) It is unlikely that these estimates of diet composition have representative value for PWS as a whole, due to the low sample size and the subsequent extrapolation to percent biomass.   ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     45dytes), smelts (Osmeridae) and euphau-siids.  Their diet also includes some 27 other fish species and 13 varieties of in-vertebrates, many of which are commer-cially important (Hart  1980).  Dogfish are known predators of juvenile Pacific salmon (Orsi et al. 1998).  Temporal pat-terns of residence for spiny dogfish in PWS are unknown. Pacific sleeper sharks are a large demer-sal species that average ~200cm in PWS (Hulbert 1999 unpublished data).  Very little is known of sleeper shark life his-tory.  Age and size at maturity are un-known.  They are thought to be ovovi-viparous, but gestation time and litter size have not been documented.   Sleeper sharks are said to be voracious and ver-satile feeders of fish. Principal prey include flatfish - halibut, soles and other flatfishes - salmon, and rockfish.  Other foods include octopods and squids, crabs, and carrion (Hart 1980).  Sleeper shark diet has also been shown to include marine mammals, including harbor seal, Phoca vitulina (Bright 1959), and southern right whale dolphin, Lissodelphis peronii (Crovetto et al. 1992). Table 53 presents a sleeper shark diet com-position based on stomach contents from the Gulf of Alaska. Temporal patterns of resi-dence for sleeper sharks in PWS are not documented.  ADF&G longline sablefish sur-veys (May and October) catch an occasional sleeper shark, but it is unknown whether they remain in the sound during winter (Robert Bercelli, Alaska Dept. of Fish and Game, pers. comm.).  Prey collected from Pacific sleeper shark stomachs in PWS during the ADF&G longline sablefish survey in September 1999 included adult coho and pink salmon (Hulbert unpub-lished data).  The sharks were caught at depths ranging from 350-550 m.  Adult pink and coho salmon depth data recorded on data stor-Table 56. Spiny dogfish prey composi-tion (% weight) sampled off the coast of British Columbia (Jones and Geen 1977). Spiny dogfish prey % in diet Unidentified teleosts 17.56 Herring (Clupeidae) 14.42 Euphausiid (Euphausiacea) 12.87 Plankton 9.09 Shrimp (Pandalus sp.) 7.57 Crab (T. Brachyura) 6.68 Gadid fish (Gadidae) 5.37 Flatfish (Pleuronectidae) 3.89 Eulachon (Osmeridae) 3.65 Octopus (Octopus sp.) 2.87 Combjellies (Ctenophora) 2.26 Elasmobranchs 1.99 Squid (Loligo sp.) 1.61 Jellyfish (Coelenterata) 1.15 Sandlance (Ammodytidae) 1.11 Rockfish (Sebastes) 0.98 Table 57. Generalized shark diet composition estimates for PWS. Percent biomass contribu-tion to shark diet are averages weighted by relative biomasses of three shark speciesa.  Generalized "shark" prey Biomass (t⋅year-1)b % in diet Adult Salmonids  1,437 13 Adult Sablefish  1,294 12 Adult gadid fish  161 1 Juvenile gadid fish  193 2 Walleye pollock  188 2 Rockfish  145 1 Herring  533 5 Sandlance  40 0 Smelt  131 1 Halibut  395 4 Flathead sole  35 0 Arrowtooth flounder  2,415 22 Other flatfish  171 2 Teleostei  12 0 Spiny dogfish  252 2 Elasmobranch fish 72 1 Squid  120 1 Octopus  272 2 Euphausiids 462 4 Combjellies  81 1 Jellyfish  41 0 Plankton 327 3 Crab  240 2 Shrimp (Pandalus) 272 3 Misc. benthic inverts 416 4 Fishery offal 441 4 Unidentified teleost 643 6 a) Generalized shark diet composition (salmon shark, sleeper shark, and spiny dogfish) is based partly upon conjecture. Salmon shark estimates were based on just 11 stomach samples from summer, and are thus unlikely to represent an annual diet. Future changes in relative abundance of the shark species, and diet composition, would act to compound uncertainty. b) Based on an estimated daily ration of 1%  c) BW⋅day-1 46     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      age tags during July and August 1999 never exceeded 90 m (Walker, et al. 1999).  Based on this information, sleeper sharks could be making vertical foraging igrations. At high latitudes sleeper sharks are known to venture into the littoral and intertidal zones and oc-casionally come to the surface (Hart 1973).  Sleeper shark diet has also been shown to include marine mammals, including harbor seal, Phoca vitulina (Bright 1959), and southern right whale dolphin, Lissodelphis peronii (Crovetto et al.  1992).  Seals are considered common prey of the Greenland shark, Somniosus microcephalus, the Atlan-tic congener of the Pacific sleeper shark (Compagno 1984).  The behavior of these species can be expected to be very similar, and sleeper sharks may prey on marine mammals in PWS (Bruce Wing 1999 pers. comm.). The consumption / biomass (Q/B) estimate for the ‘PWS sharks’ group in early versions of the model was based on a conservative daily ration estimate of 1% per day. The Q/B estimate was then corrected for PWS residency time for each species, and the re-sulting species-specific Q/B estimates were averaged (weighted by PWS biomass).  New information indicating that salmon sharks may consume 5% of their body weight per day rather than the initial as-sumption of 1%. This corresponds with a Q/B of 18.3 year-1 and to a residency time-corrected Q/B of 9.1 year-1 for that species , which in turn leads to a Q/B value of 7 year-1 for overall PWS sharks by using a bio-mass-weighted average, also corrected for residency time (Table 54). This adjustment was based on an increase in daily ration esti-mates from 1% to 5% in salmon sharks due to metabolic considerations. Table 57 presents a generalized shark diet composition by combining the estimated an-nual diet compositions for all three species of sharks, weighted by proportions of biomass represented by each species.   BIRDS Invertebrate-Eating Sea Ducks Dan Esler Alaska Biological Science Center, USGS Anchorage, Alaska, USA This group is comprised of the primary ben-thic invertebrate-eating sea ducks that occur in Prince William Sound (i.e., 8 species listed in Table 58). Excluded from this analysis are rare sea ducks (e.g., eiders) and fish-eating sea ducks (i.e., mergansers). Data summarized in Table 58 were used to calculate an annual mean biomass of 0.005 t⋅km-2 and a Q/B ratio of 450.5 year-1. The ratio of production to biomass (P/B) was considered to be equal to 0.2 year-1, the estimated mortality for each of these species.  Population estimates are from the most recent U.S. Fish and Wildlife Service (Migratory Bird Management) surveys (Agler and Kend-all 1997).  Table 58. Ecopath parameters for invertebrate-eating sea ducks in PWS Species Winter popu-lation Summer population Body weight (kg) Winter biomass  (t) Summer biomass (t) Prey  consumed (kg⋅bird-1⋅day-1) Food consumption (t⋅year-1) Harlequin Duck 17,151 10,619 0.60 10.29 6.37 0.66 3,472 Goldeneyes 35891 0 0.90 32.30 0 0.99 7,532 Surf Scoters 6492 3024 1.10 7.14 3.33 1.21 2,225 White-winged Scoters 6203 0 1.35 8.37 0 1.49 1,952 Black Scoters 1837 0 1.15 2.11 0 1.27 492 Oldsquaw 6852 0 0.90 6.17 0 0.99 1,438 Bufflehead 6875 0 0.45 3.09 0 0.50 721 Total 81,301 13,643 -- 69.48 9.70 -- 17,835 ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     47 Biomass estimates are simply the number of individuals multiplied by average body weight. Note that both numbers and biomass are seasonally variable. Prey weight consumed per day was esti-mated based on relationships described for Barrow’s goldeneyes (Holland-Bartels 1997). Goldeneye field metabolic rate (FMR) was estimated to be 1674kJ⋅day-1 using an equation for flapping flight sea birds (Birt-Friesen et al. 1989). Prey energy density of mussels was estimated to be 1.65kJ⋅gram-1 wet weight including shell (Palmerini and Bianchi 1994; Mary Ann Bishop, unpubl. data). Thus, the estimate of wet weight consumed per day for gold-eneyes was 1015g. Based on goldeneye av-erage body weight of 900g, individuals con-sume an average of 110% of their body weight daily. This seems high but may be accurate given the high water content and low energy density of mussels, especially when their shell is included. This figure was applied to other species, which assumes (1) no variation in FMR with body weight and (2) energy density of all prey items is similar to mussels. Prey weight consumed per day was esti-mated based on relationships described for Barrow’s goldeneyes (Holland-Bartels 1997). Goldeneye field metabolic rate (FMR) was estimated to be 1674kJ⋅day-1 using an equation for flapping flight sea birds (Birt-Friesen et al. 1989). Prey energy density of mussels was estimated to be 1.65kJ⋅gram-1 wet weight including shell (Palmerini and Bianchi 1994; Mary Ann Bishop, unpubl. data). Thus, the estimate of wet weight consumed per day for goldeneyes was 1015g. Based on goldeneye average body weight of 900g, individuals consume an aver-age of 110% of their body weight daily. This seems high but may be accurate given the high water content and low energy density of mussels, especially when their shell is in-cluded. This figure was applied to other spe-cies, which assumes (1) no variation in FMR with body weight and (2) energy density of all prey items is similar to mussels. Food consumption per year is estimated by calculating bird days (adjusted for seasonal changes in abundance) and multiplying by daily food requirements. Diet data in Table 59 are gathered from pub-lished sources (Vermeer 1981, Koehl et al. 1982, Sanger and Jones 1982, Vermeer and Bourne 1982, Vermeer 1982, Goudie and Ankney 1986, Goudie and Ryan 1991, Patten 1994). Sea duck diets vary considerably by site and few studies have been conducted in Prince William Sound; the data presented rep-resent my assimilation and summary from all available sources.   Table 59.  Diet composition (% weight) of invertebrate-eating sea ducks in PWS Species Mussels Clams Snails Chitons Crustaceans Limpets Harlequin Duck 10 0 35 10 35 10 Goldeneyes 90 -- 10 -- -- -- Surf Scoters 75 10 5 -- 5 5 White-winged Scoters 30 30 30 -- 10 -- Black Scoters 85 -- -- -- 15 -- Oldsquaw 20 20 20 -- 40 -- Bufflehead 5 5 40 0 45 5 Means 45 9.3 20 1.4 21.4 2.9 48     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Seabirds and Seabird Predators William D. Ostrand and David B. Irons US Fish and Wildlife Service Anchorage, Alaska, USA Although the seabird population of Prince William Sound is a rich and diverse collec-tion of species (Isleib and Kessel 1973) with differing foraging strategies (Klosiewski and Laing 1994, Ostrand et al. 1998), their dis-tribution is consistent across taxa, with most bird observations occurring within 1 km of the shoreline (Ostrand and Maniscalco 1996). Within the nearshore zone, seabirds have been associated with shallow water habitats. However, this relationship was not apparent during 1997 (U.S. Fish and Wildl. Serv., Anchorage, unpubl. data). Population estimates in Table 60 are from 1996 U.S. Fish and Wildlife Service surveys (Agler and Kendall 1997). Bird population estimates were based on counts of adult birds and did not include estimates of off-spring abundance. Mortality among seabird offspring is high (Ashmole 1971). However, we speculate that little of this biomass is returned to the Prince William Sound marine system. Much mortality of eggs and chicks is due to predation by avian predators (Hatch and Hatch 1990) and possibly mam-mals (Seto and Conant 1996). Avian nest predators are composed of marine [Glau-cous-Winged Gulls (Larus glaucescens)] and terrestrial [Common Ravens (Covus corax) and Northwest Crows (Corvus cauri-nus)] birds (Parrish 1995). We assume that none of the biomass consumed by terrestrial predators and only a portion of that con-sumed by marine predators is returned to the sea. We further speculate that young sea-birds leave Prince William Sound soon after fledging, hence mortality among these indi-viduals does not contribute biomass back into the system. Therefore, we have consid-ered only the adult population in this model-ling exercise. The summer population and biomass is dominated (>20,000 individuals of each species) by Glaucous-winged Gulls, Black-legged Kittiwakes (Rissa tridactyla), and Brachyrampus murrelets (Marbled Mur-relet, B. marmoratus and Kittlitz’s Murrelet, B. brevirostris). The winter population and biomass differs and is dominated by Mew Gulls (Larus canus), murres (mostly Common Murre, Uria aalge), and Brachyrampus mur-relets. In addition, Bald Eagles (Haliaeetus leucocephalus) are a major contributor to avian biomass (>1.5 kg km-2) during both sea-sons.  Body weight estimates for Alcids were taken from De Santo and Nelson (1995), cormorant estimates from Johnsgard (1993), and all other species from Dunning (1993). Daily food con-sumption estimates for Black-legged Kitti-wakes and Pigeon Guillemot (Cepphus columba) were obtained from studies con-ducted in Prince William Sound (U.S. Fish and Wildl. Serv., Anchorage, unpubl. data). Bald Eagle and Peregrine Falcon (Falco pere-grinus) consumption estimates were obtained from Stalmaster and Gessaman (1984) and Nelson (1977), respectively. For all other spe-cies daily food consumption was calculated using the following formula of Birt-Friesen et al. (1989): where energy is expressed in kJ and body weight in kg. We assumed a 75% efficiency in converting energy consumed and a local en-ergy content of 4.5 kJ gm-1 of forage fish (D. Roby, Oregon State Univ, Corvallis, pers. comm.). Hence, we divided daily energy by 0.75 and then divided that product by 4.5 kJ gm-1 to obtain daily consumption in wet weight.  Food habits for Pigeon Guillemot (Cepphus columba), Marbled Murrelets, Black-legged Kittiwakes, Glaucous-winged Gulls, and Mew Gulls (Table 61) were obtained from local studies (U.S. Fish and Wildl. Serv., Anchor-age, unpubl. data). Tufted Puffin (Fratercula cirrhata) data were also collected in Prince William Sound (Piatt et al. 1998). Bald Eagle, Peregrine Falcon, and cormorant food habits log10(daily energy) = 3.08 + 0.667 log10(body weight) ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     49 were taken from Grubb and Hensel (1978), Nelson (1977), and Robertson (1974), re-spectively. All other food habits data were obtained from Sanger (1987).  Estimated production biomass ratio (P/B) values of 0.078 year-1 for piscivorous sea-birds and 0.05 year-1 for seabird predators was determined by calculating the average adult mortality, weighted by species bio-mass. Adult mortalities for Black-legged Kittiwakes (U.S. Fish and Wildl. Serv., An-chorage, unpubl. data) and Bald Eagles (Bowman et al. 1993) were obtained from local studies. Mortality values for Tufted Puffins, Horned Puffins (Fratercula corni-culata), and Parakeet Auklet (Cyclorrhyn-chus paittacula) were not available so we used an Atlantic Puffin (Fratercula arctica) value (del Hoyo et al. 1996). Similarly, Her-ring Gull (Larus argentatus) mortality (Ashmole 1971) was used for Mew Gulls. Mortality values for Fulmars and Shear-waters, Marbled Murrelets, Ancient Mur-relets (Synthliboramphus antiquus), Glau-cous-winged Gulls, all cormorants, Arctic Terns (Sterna paradisaea), and Peregrine Falcons were obtained from Ashmole (1971), Beissinger (1995), De Santo and Nelson (1995), Reid (1987), Johngard (1993), Coulson and Horobin (1976) and Ambrose and Riddle (1988), respectively. 50     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      50 Table 60.  Population estimates and Ecopath input parameter estimates for seabirds in Prince William Sound. Species Summer populationa 95% CI Winter populationa 95% CI Body weight (g) Summer biomass (kg⋅km-2) Winter biomass (kg⋅km-2) P/B (year-1) Prey weight (g⋅day-1 ) Q/B (year-1) Food consumption (kg⋅km-2⋅year-1) Seabirds            Fulmars and Shearwaters 18779580 0 8100.2 0.0 6.0 309.5139.5 12.8 Fork-tailed Storm-petrel 15800114510 -- 430.1 0.0 -- 43.7370.7 14.3 Double-crested Cormorant 74110367 230 23500.0 0.1 15.1 629.897.8 5.8 Pelagic Cormorant 263225590 552 20000.1 0.1 17.0 565.6109.5 10.0 Unid.: red-faced or pelagic 1067150812056 4005 20000.2 2.7 17.0 565.6109.5 153.9 Bonaparte's Gull 160013430 0 2120.0 0.0 15.0 126.6217.9 4.2 Mew Gull 14200552620300 11702 4000.6 0.9 15.0 193.3176.4 138.4 Glaucous-winged Gull 25100654713900 5442 10102.9 1.6 15.0 358.6129.6 289.8 Black-legged Kittiwake 48227188825279 2129 3902.2 0.3 7.0 190.1177.9 218.3 All terns 540017100 0 1100.1 0.0 13.0 81.7271.2 9.1 Arctic Tern 485216560 0 1100.1 0.0 13.0 81.7271.2 8.2 All Murres 3300217746100 19571 10040.4 5.3 11.0 357.2129.8 366.8 Pigeon Guillemot 29829052500 1056 4870.2 0.1 20.0 220.4165.2 25.1 Brachyramphus murrelet 822001891744300 13158 2212.1 1.1 15.0 130.1214.9 341.2 Ancient Murrelet 1881850 0 2060.0 0.0 23.0 124.2220.0 0.5 Parakeet Auklet 8004190 0 2970.0 0.0 5.0 158.5194.8 2.6 Tufted Puffin 500021260 0 7730.4 0.0 5.0 300.0141.7 31.0 Horned Puffin 5003900 0 6120.0 0.0 5.0 256.7153.1 2.7 Sum (mean) 213430--145392 -- --9.5 12.3 (7.8) --(150.6) 1634.7         Seabird Predators         Bald Eagle 30467413893 832 47001.6 2.1 5.0 489.036.5 70.0 Peregrine Falcon 670 0 11300.0 0.0 23.0 150.048.5 18.0 Sum (mean) 3052--3893 -- --1.6 2.1 (5.0) --(38.9) 88.0 a) Summer: May-October; Winter: October-April. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                      5151 Table 61. Diet composition (% weight) of Seabirds in Prince William Sound Predators\Prey Adult salmon Salmon fry 6-12cm Inshore detritus Sand- lance Juv. herring Capelin Nearshore demersals Pollock 0 Pollock 1-3 Squid Herb zoo. Omni. zoo. Shal. sm.epiben. Sea-birds Other marine Non- marine Seabirds           Fork-tailed Storm Petrel - - -- - -- 5- - -- - Sooty Shearwater - - -5.0 - 70- -- 22.0-3.0 - -- - Double-Crested Cormorant - - -5.0 3.0 -92.0 -- --- - -- - Pelagic Cormorant - - -18.9 - -74.3 -- --6.8 - -- - Mew Gull - - 40.020.0 20.0 20- -- --- - -- - Glaucous-winged Gull - - 40.012.0 12.0 6- -- --1.2 30.0 -- - Black-legged Kittiwake - 4.0 3.019.0 61.0 12- -- --5.0 - -- - Arctic Tern - - -1.0 1.0 1- -- --97.0 - -- - Common Murre - - -20.0 - 40- -25.0 --15.0 - -- - Pigeon Guillemot - - -17.4 3.9 -62.4 -11.8 --- - -4.5 - Marbled Murrelet - - -43.0 48.0 -- -- --- - -9.0 - Ancient Murrelet - - -- - -- -13.0 2.0-- - -5.0 - Tufted Puffin - 24.0 -- 22.0 -- -13.0 --- - -41.0 - Horned Puffin - - -18.0 - 65- -1.0 2.0-- - -12.0 - Mean - 1.0 10.822.0 22.4 148.7 06.0 5.3 -2.7 -            Seabird predators            Bald Eagle 10.0 - 10.0- - -5.0 -- --2.5 - 3010.0 25 Peregrine Falcon - - -- - -- -- --- - 100- - Mean 8.0 - 8.0- - -4.0 -- --2.0 - 508.0 20 52     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Consumption of Herring Eggs by Birds Mary Anne Bishop Pacific NW Research Station, USFS  Cordova, Alaska, USA Thomas A. Okey Fisheries Centre, UBC, Canada Consumption of herring eggs is incorporated into bird diet compositions in the previous two sections at an ‘annual’ resolution (as ‘inshore detritus’ for the purposes of the model). Additional information on this phe-nomenon is provided in this section to document the smaller spatial and temporal resolutions at which some energy flows oc-cur. Herring eggs are deposited on kelp and along the shoreline at Montague Island and northeast PWS during a few weeks in the spring. Herring eggs are thus an ephemeral resource for migratory birds in PWS, but these deposited eggs can be a substantial food source, particularly considering their high energy content. The estimated biomass of herring eggs in PWS was 1,413 t in 1995 and 1,484 t in 1996, with a mean of 1,449 t, or 0.160 t⋅km-2 when expressed on a Sound wide basis. However, the 1990 biomass es-timate was almost nine times greater than the 1995-1996 period, at 12,826 t (J. Wil-cock, ADF&G, unpublished data). At northern Montague Island in 1994 and 1995, information on the abundance and distribution of the five most numerous avian herring spawn predators was collected using boat and aerial surveys. These species in-cluded Glaucous-winged Gulls, Mew Gulls, Surf Scoters, Surfbirds, and Black Turn-stones. In 1995, problems with aerial video-graphy prevented an estimate of Mew Gull and Glaucous-winged Gull abundance. For each species, a daily herring spawn consump-tion per individual bird was determined using a bioenergetic model based on field metabolic rate, energy content of spawn, and proportion of energy acquired from herring spawn. En-ergy acquired from herring spawn was deter-mined based on stomach content analyses of birds collected in Montague Island spawn ar-eas during 1994 and 1995. Glaucous-winged Gulls, Surf Scoters, and Mew Gulls were con-suming only spawn (Table 62). Intake of her-ring eggs by birds ranged from 1.06 kg per Surf Scoter per day to 0.16 kg per Black Turnstone per day, with other birds consum-ing intermediate quantities (Table 63). Total herring spawn consumption in 1994 at the northern Montague Island study area was es-timated to be 729 t. As part of a separate EVOS study, in 1997, Glaucous-winged Gulls were collected in win-ter and spring at northern Montague Island, from Stockdale Harbor to Port Chalmers. Spring gull collection was conducted during four time periods: prior to spawn, active spawn deposition, spawn incubation, and post spawn hatch.  Glaucous-winged Gulls con-sumed adult herring prior to and during active spawn deposition, switching to spawn once deposition was complete in an area (Table 64). In 1997, spawn covered many other areas at Montague, therefore, estimates of gull numbers for all spawn areas are not available.  ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                      53 53 Table 62.  Percent occurrence (% occ.) and aggregate % weight (% wt.) of prey items contained in gut samples of birds collected from within spawn areas on northern Montague Island, April-May 1994, 1995.      Glaucous-winged Gull (n=13)a    Mew Gull (n=9)a  Surf Scoter (n=8)a Surfbird (n=20) Black Turnstone (n=14)                                  Species  % occ.b   %wt.c   % occ.  %wt.    % occ.  %wt.    % occ.  %wt.  % occ.  %wt.                                   Fish                                                         Herring Egg 100   100      100    96      100    100      75    70.5    69    74.0            Bivalves                                                          Mytilus  -      -        -      -        -      -        80    27.7    19    1.6            Crustaceans                                                          Balanus  -      -        -      -        -      -        5    0.2    19    23.6              Amphipod sp.  -      -        -      -        -      -        -      -      6    0.9              Amphithoe sp.  -      -        -      -        -      -        5    <0.1    -      -                Hermit crab  -      -        -      -        -      -        10    0.3    -      -              Gastropods                                                          Alia sp.  -      -        -      -        -      -        10    0.7    -      -                Lirularia sp.  -      -        -      -        -      -        5    <0.1    -      -                Littorina sitkana  -      -        -      -        -      -        10    0.1    -      -                Margarites sp.  -      -        -      -        -      -        10    0.4    -      -                Unid. gastropod  -      -        -      -        -      -        5    <0.1    -      -              Insects                                                          Diptera Larvae  -      -        11    4.2      -      -        -      -      -      -              Nematods                                                          Nematode  8    <0.1       -      -        -      -        -      -      -      -              Unid. organic material  8    <0.1       -      -        -      -        -      -        -      -         a  Aggregate weight based on 12 glaucous-winged gulls, 8 mew gull, and 7 surf scoters;   b  Percent occurrence: number of individuals with prey item x100 / total number of individuals; c  Percent aggregate weight = total weight of prey item for all individuals x100 / total weight of all prey items for all individuals. 54     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Herring eggs are not explicitly defined as a group in the PWS Ecopath model at the present time for the sake of simplic-ity. Consumption of herring eggs by modeled avian groups (i.e., Mew Gulls, Surf Scoters, Glaucous-winged Gulls), is represented as consumption of inshore detritus (Ostrand and Irons, this vol.; Table 61). Avian herring spawn preda-tors (Black Turnstones, Surfbirds) are sufficiently transitory that their exclu-sion from the model is somewhat justi-fied. However, such transitory species are part of the PWS ecosystem, and in the future, herring eggs may be consid-ered as an explicit group and their transi-tory predators explicitly included.  Table 63. Daily herring spawn consumption by avian species based on bioenergetic model. Northern Montague Island 19 April - 15 May 1994 and 27 April - 19 May 1995. Species Mean weight (kg)/bird Aggregate energy for eggs (%) Eggs consumed per day (kg) Total bird days (1994) 1994 spawn consumed (t) Total bird days (1995) 1995 spawn consumed (t) Glaucous-winged 1.33 100 0.73 825,156 601.1 -- -- Mew Gull 0.45 95 0.32 213,755 67.8 -- -- Surf Scoter 1.16 100 1.06 42,392 44.9 24,558 26.0 Surfbird 0.21 93 0.19 73,742 14.3 102,248 19.8 Black Turnstone 0.14 99 0.16 6,297 1.0 12,432 1.9 Table 64. Percent occurrence and percent aggregate weight of prey items in gut samples (esophagus/proventriculus) of glau-cous-winged gulls. Northern Montague Island, 13 Dec 96-24 Feb 97 (winter), and 14 Apr 97-12 May 97 (spring). Spring (n = 30) Species Winter  (n = 10) Prespawn (n = 8) Spawn (n = 4) Postspawn/  Prehatch (n = 14)  % Occa % Wtb % Occ % Wt % Occ % Wt % Occ % Wt Eggs: Herring Eggs - - - - 25 5.9 100 92.4 Fish: Clupeas Pallasi - - 37.5 69.4 - - - -    Offal 10 4.1 - - - - - -    Unidentified Fish - - 25 28.6 100 94.1 7.1 7.6 Stars: Evasterias trochelli 20 69.0 - - - - - -    Pycnopodia helianthoides 30 23.0 - - - - - -    Pisaster ochraceus 10 3.3 - - - - - -    Unidentified Sea Star 10 0.4 12.5 2.0 - - - - Crustaceans:  Cancer sp. 10 0.03 - - - - - - Unidentified Organic Matter 20 0.2 - - - - - - a) Percent Occurrence: number of birds with prey type / total birds b) Percent Aggregate Weight: weight of prey type / total weight of all prey items ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     55MAMMALS Baleen Whales Craig Matkin North Gulf Oceanic Society Homer, Alaska, USA Rod Hobbs Alaska Fisheries Science Centre NOAA, Seattle, Washington, USA Baleen whales are represented by humpback whales, for which however, we have no re-cent population estimates for Prince William Sound. Waite et al. (in press) use abun-dances of 140 to 200 to represent a range for the population in Prince William Sound, the south side of the Kenai Peninsula and adja-cent waters. The range found during the post-EVOS studies in 1989 and 1990 proba-bly still holds (O. von Ziegesar et al. 1994). Total numbers that use the study area of the model in a given year probably range from 60-90 whales (O. von Ziegesar, unpub. data). Some of these stay all winter, or at least there are humpback whales in the Sound during all months of the winter. There are probably the fewest between mid-January to mid March, but there may be as many as 10-20 even during these months (O. von Ziegesar, unpub. data). We have done no winter surveys, but observations from winter herring surveys indicate humpback whales are associated with the herring all fall and winter. The amount these whales eat is estimated at 30 g⋅kg-1⋅day-1. and an aver-age humpback is estimated to weigh 32.7 t (from Dolphin 1987). In the summer a ma-jority of prey are euphausiids based on sonar scans (M. Nerini, formerly with NMFS Na-tional Marine Mammal Lab., unpub. data) and scats observed opportunistically (O. von Ziegesar, unpub. data), although they cer-tainly feed on sand lance and herring as well. It appears that the diet shifts toward herring in the fall (late September, October) and into winter, but this is from observational data from herring researchers in late fall and winter. Estimated population parameters for baleen whales are listed, along with those for other whales, in Table 65.   Sea otter  James L. Bodkin, Dan H. Monson and George E. Esslinger Alaska Biological Science Center, USGS—BRD Anchorage, Alaska, USA The purpose of this section is to provide esti-mates of biomass, mortality, prey consumption and dietary composition for sea otters (Enhydra lutris) in Prince William Sound, Alaska. Bio-mass estimates are derived from an aerial sur-vey conducted in 1994. Sea otters in Prince William Sound are distributed by bathymetric contours as follows; shoreline to 40 m contour = 0.85. 40-100 m contour = 0.10 and >100m = 0.05 (Bodkin and Udevitz 1996). Prey data were obtained from field observations in 1996 and 1997. We assume biomass, mortality, and prey composition remain constant over time. We assume sea otters require 25 % of their weight/day in prey (not including shells) (Costa 1982), and used conversions from Kvitek (1992) to estimate wet weight of clams from shell length.  Food habits data were extracted from Holland-Bartels (1997) and unpublished data from the authors. Estimates of sea otter population parameters are found in Table 66. Table 65. Cetacean input parameters for PWS Group Biomass (t⋅km-2) +/-  (t) P/B  (year-1) Q/B (year-1) Baleen whales 0.1486 0.0743 0.05 10.95 Small cetaceans 0.0088 0.0044 0.10 29.20 Transient orcas 0.0019 0.0013 0.05 6.00 Resident orcas 0.0113 0.0075 0.05 8.67 Aggregated orcas 0.0132 -- 0.05 8.29 a) P/B and Q/B estimates for PWS were derived by multiplying annual estimates for the group by the estimated fraction of a year spent in PWS (0.25).  Table 66.  Estimates of sea otter population parameters for PWS. Biomassa (t⋅km-2) 95% CI  (±; t⋅km-2) P/Bb (year-1) Range (year-1) Q/Bc (year-1) Range (year-1) 0.0450 0.0150 0.13 0.10-0.15 117 100-140 a) Biomass derived from 1994 survey of PWS including Orca inlet in which the population size was estimated at 14,352 individuals (see above tables); b) Corresponds to instantaneous rate of total mortality; c) Ration for an average sized individual. 56     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Methods We conducted a survey throughout all of Prince William Sound, including Orca Inlet, in July of 1994. The aerial sea otter survey methodology consists of two components:  (1) strip and (2) intensive search units (Bod-kin and Udevitz 1991, 1996). Sea otter habi-tat was sampled in two strata, high density (shoreline - 40 m contour) and low density (40 - 100m contour), distinguished by dis-tance from shore and depth contour. Survey effort was allocated in proportion to ex-pected sea otter abundance (0.85 for high abundances, and 0.15 for low abundances) by adjusting the systematic spacing of tran-sects within each stratum. Transects with a 400 meter strip width on one side of a fixed-wing aircraft were surveyed by a single ob-server at an airspeed of 65 mph (29 m⋅s-1) and altitude of 300 feet (91 m). The observer searched forward as far as conditions allow and out 400 m, indicated by marks on the aircraft struts, and recorded otter group size and location on a transect map. A group was defined as one or more otters spaced less than three otter lengths apart. Intensive search units (ISUs) were used to estimate the proportion of sea otters not detected on strip transect counts. ISUs were flown at intervals dependent on sampling intensity throughout the survey period, and were ini-tiated by the sighting of a group, then fol-lowed by five concentric circles flown within the 400-m strip perpendicular to the group that initiated the ISU. Two observers were used in 1994, resulting in separate es-timates of detection for each observer. Food habits and foraging success of near-shore feeding sea otters were measured dur-ing shore based observations in 1996 and 1997. High power telescopes (Questar Corp., New Hope, PA) and 10X binoculars were used to identify prey type (lowest pos-sible taxon), prey number, and prey size (small <5 cm, medium 5-9 cm, and large >9 cm), and dive success (prey captured or not) during foraging ‘bouts’. A ‘bout’ consisted of observations of repeated dives for a focal ani-mal while it remained in view and continued to forage (Calkins 1978). Assuming each foraging bout records the feeding activity of a unique individual, bouts were considered independent while dives within bouts were not. Thus the length of any one foraging bout was limited to one hour after which a new focal animal was chosen. Results In July 1994, we conducted an aerial survey of sea otters in PWS (Bodkin and Udevitz in press), which included 7,328 km2, of which 2,987 km2 were considered high density stratum and 4,341 km2 low density stratum. The results of the survey are presented in Table 67. It is likely that the estimate of abundance generated from this survey methodology are negatively biased by about 5-10%, due to detection prob-abilities of 90-96% during survey development experiments (Bodkin and Udevitz 1991). Al-though there may be small scale (10s of km) movements of sea otters seasonally, we believe that the overall abundance of sea otters in Prince William Sound does not vary seasonally.  We calculated a mean sea otter weight of 23.0 kg based on actual weight of >116 sea otters captured and weighed in western PWS in 1996 and 1997. This includes 79 females and 37 males, roughly in proportion to the sex ratio in the population. We estimated an instantaneous mortality rate of 0.13 year-1 based on an average age of 7 years in the live population. We as-sumed that immigration approximates emigra-tion.  A total of 1425 foraging dives were observed Table 67.  Otter counts, unadjusted population size estimates and adjusted population size es-timate in the 1994 sea otter survey, PWS. Type of estimate Population estimate (N)  s.e. Unadjusteda 1085 2051 Adjustedb 14352 2418 a) 1973 otters were observed on the 681 km2 transects; b) See correction factors (Table 68). Table 68. Correction factors for population estimates Observer No. of ISUs Factor S.E. J.B. 42 1.92 0.20 G.E. 55 1.39 0.08 ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     57 between 19 May and 23 July, 1996, including 631 dives dur-ing 70 forage bouts at Knight Island (mean of 9 dives/bout), and 794 dives during 58 bouts at Montague (mean of 13.7 dives/bout).  A total of 1,271 foraging dives were observed between 23 May and 15 August, 1997, in-cluding 604 dives during 49 forage bouts at Knight Island (mean of 12 dives/bout), and 667 dives during 58 bouts at Montague (mean of 11.5 dives/bout).  Prey composition was similar between areas and was dominated by clams. Butter clams (Saxidomas giganteus) were the most com-monly identified species followed by soft shell clams.  Prey types other than clams included small crabs (primarily Telmessus), fat innkeepers (Echiurus), and sea stars (Ev-asterias and Leptasterias) all of which were uncommon or missing from the diet at Knight. Sea urchins (Stongylocentrotus droebachiensis) were rare in the diet in both areas, though urchins were present in scats examined in both areas (4 of 44 examined at Knight (9%), and 6 of 43 examined at Mon-tague (14%), possibly reflecting seasonal differences in sea urchin utilization.  Prey composition for the purpose of this re-port is summarized by faunal category in Table 69 and is a compilation of food habit data collected during the summer of 1996 and 1997. It is likely that prey composition varies geographically within the Sound, but our results are similar to results of others (Calkins 1978, Doroff and Bodkin 1994).  It is also possible that seasonal differences in prey species composition exist but we have no data to address this possibility   Pinnipeds Kathy Frost Alaska Dept. of Fish and Game Fairbanks, Alaska, USA This group consists of harbor seals (Phoca vi-tulina richardsi) and Steller sea lions (Eumeto-pias jubatus). Harbor seals occur in most coastal areas throughout PWS, particularly Northwestern, Southwestern, and Eastern areas. Sea lions are far less abundant than harbor seals in the PWS as defined here. Harbor seals occur in PWS throughout the year. Data from 50 seals satellite tagged during 1992-1997 indicate that most PWS harbor seals show strong site fidelity, remaining near the haulouts where they were originally tagged. Some seals make feeding trips to the Gulf of Alaska (GoA), the Copper River Delta, or between the northern and southern sound, especially during fall through spring. Usually these seals return to PWS during their feeding trips. Occasionally longer movements are made (to Cook Inlet or Yakutat) and seals may or may not return to PWS.  Twelve newly-weaned harbor seal pups tagged in 1997 also made relatively local movements, with occasional trips to the GoA by a few. Since satellite tags remain attached only until the following molt (usually 9-11 months after tagging), the tags do not provide informa-tion about any long-term movements that might occur. However, several tagged seals have been recaptured 1-3 years later near the original cap-ture location. Based on GoA samples collected in the 1970s, Table 69. Diet matrix for sea otters in PWSa. Taxa % in diet PWS model component % in diet Clams 80 Shallow large infauna 40 Mussels 12 Deep large infauna 40 Crabs 4 Shallow large epifauna 16 Other 4 Deep large epifauna 4 a) Fraction refers to weight, or volume, of energy units (NOT frequency of occurrence) Table 70. Ecopath input parameters for harbor seals in PWS, 1992-1997 Species  Biomass (t⋅km-2) P/B (year-1) Q/B (year-1) Export (t⋅km-2⋅yr-1) Harbor Seal 0.05 (± 0.04-0.06) a 0.06 25.55b 0.002 a) Based on the range of population estimates since 1990 (does not represent 95% CI); b) Converted by the editors from Q = 2,044 kg⋅seal-1⋅year-1 in PWS  58     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      cumulative mortality of harbor seals from birth to 4 years was estimated at 77% (Pitcher and Calkins 1979). At 4-7 years, mortality was 9-11% per year. Mortality remained fairly constant at 9-10% per year until age 20, increasing to about 14% after age 20. Mortality may be substantially dif-ferent in other areas. For example, in British Columbia where the harbor seal population has been rapidly growing, subadult mortality was about half what it was in the GoA (Bigg 1969). There are no recent data for the GoA (including PWS). Maximum recorded age for harbor seals in Alaska is 32 years. In the GoA, sex ratio is about 50:50 until seals reach 20 years of age. Then, the ratio is ap-proximately 78% females. Table 70 shows estimated population pa-rameters for harbor seals in PWS. The esti-mated population size of harbor seals in PWS is 5,500 (range 4,600 - 6,400). This value is the mean (range) of Trend A + Trend B x 1.61 correction factor. It does not include every seal in PWS, and no confi-dence intervals are available. The average size of adults in the GoA during the mid 1970s was 84.6 kg for adult males and 76.5 kg for adult females. Assuming a 50:50 population of males and females, and a PWS population size of 5,500, there would be 443 t of harbor seal in PWS. Adding 10% for Steller sea lions gives about 487 t of pin-nipeds in PWS, or 0.054 t⋅km-2. Daily consumption by harbor seals ranges from 6-8% of body weight per day (Ash-well-Erickson 1981; Ashwell-Erickson and Elsner 1981), and depends on the caloric content of the prey and season. Captive feeding experiments showed that harbor seals consume ~4% daily of their body weight in March through August and ~8% in winter. About 40% of the total annual net energy required by the harbor seal popula-tion is necessary to sustain 0-3 year olds (50% of the energy goes to growth from birth to weaning stages, compared with 2-7% from weaning to three years, and less than 2% from 2-24 years). Values given in Table 70 for PWS harbor seal P/B, Q/B, and export are used here for PWS pinnipeds. Harbor seal diet varies by age and season. There are no recent stomach contents data for PWS harbor seals. Based on data from the mid 1970s, in order of descending frequency of occurrence, pups ate capelin, pollock, tomcod, and cephalo-pods; yearlings ate herring, pollock, squid, cap-elin, eulachon; other subadults ate pollock, her-ring, tomcod, cod, capelin, flatfish, squid; and adults ate pollock, herring, cod, eulachon, octo-pus, squid, tomcod, flatfish, and salmon. By month the most commonly eaten prey were: February - pollock, herring, cod, and cephalo-pods; March  - pollock, capelin, herring, and cod; April - herring, pollock (capelin, eula-chon); May/June - eulachon and pollock; July - eulachon, herring, pollock, and tomcod; Sep-tember/October - pollock, tomcod, herring, cod, and flatfish; November - pollock, squid, octo-pus, and cod. Flatfishes were also eaten, but the amount is unknown. Recent data from analysis of fatty acid signatures in blubber indicates that these same species were still present in harbor seal diets in the mid 1990s. Additional analyti-cal models must be developed before the rela-tive importance by species can be estimated for the fatty acids data. An estimated pinniped diet composition is shown in Table 71. The kill of harbor seals is about 250 year-1, or 0.002 t⋅km-2⋅year-1. Steller sea lions are not hunted.  Table 71. Pinniped diet composition (% weight) in PWS, 1992-1997 Prey Harbor seals Pinnipedsa Pollock 47 52 Herring 11 12 Squid 6 12 Salmon 10 10 Capelin 4 5 Nearshore pelagics 5 -- Octopus 5 -- Shallow Small Epifauna -- 5 Eulachon 2 2 Pacific Cod 1 2 other 9 -- a)  Adapted from harbor seal diet by R. Hobbs (including Steller sea lions) ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     59 Orcas Craig Matkin North Gulf Oceanic Society Homer, Alaska, USA Rod Hobbs Alaska Fisheries Science Centre NOAA, Seattle, Washington, USA Transient Orcas There are 11 transient killer whales (in group AT1) that seem to spend most of their time in the PWS ‘region’ and other tran-sients that come and go from other regions in the Gulf of Alaska. About 8 of those are in the Prince William Sound region at any given time (C. Matkin unpublished data). We do not know how these numbers change in the wintertime. For the purpose of this exercise we assume that average residence time in the study area for the 19 whales is 3 months with a range from 1 to 6 months. The average weight of a killer whale is 3550 kg (Baird 1994). Barrett Lennard et al. (1995) has the best estimates of killer whale food consumption; for transients, this was 58.7 kg of marine mammal⋅day-1 (lower for transients than residents because of higher caloric values for pinniped and small ceta-cean prey). A 25% correction was added (for wild/captive diet extrapolation) for tran-sients and residents, though this is question-able (C. Matkin; and Barrett Lennard, pers. comm). From observations of transient killer whales the diet consisted of 32% harbor seal, 39% Dall’s porpoise, 6% harbor por-poise, and 23% unidentified marine mam-mals. (Saulitis et al., unpublished data). For the purpose of this study we have prorated the unidentified marine mammals into the three identified species. Resident Orcas  There are 112 resident type killer whales that center their range in PWS. This center seems to have recently shifted a bit west-ward, into the Kenai Fjords region. For the purpose of this exercise we assume that av-erage residence time in the study area for the 112 whales is 3 months with a range from 1 to 6 months. Barrett-Lennard et al (1995) has the best estimations of killer whale food con-sumption. The rate for resident killer whales fish consumption was estimated at 84.3kg of fish⋅day-1 (176K kcal⋅day-1). A 25% correction factor was applied here to increase this daily consumption figure for both types of killer whales but one of us (C. Matkin) thinks that this adjustment is questionable. The average weight of a killer whale is 3550 kg (Baird 1994). Also from this study we found 95% of the scale sam-ples collected from resident fish kills were coho salmon, the rest being divided between chum and chinook salmon (Saulitis et al., unpublished data). Although there may be some bias in the sampling, the dependence on specific salmon species, particularly coho salmon is probably real. Observations indicate that resident killer whales begin feeding on herring in April. We have no way of knowing how the diet changes for either residents or transients during the No-vember-March period.  Orca aggregation and disaggregation (Thomas A. Okey) The biomass used for the aggregated orca group (0.003) is ¼ of the sum of the biomasses of each orca group because the average residence time in PWS is 3 months. The Q/B and the P/B values were derived by calculating means weighted by the relative biomass of each orca group presented above (Table 65). Aggregated diet composition was likewise calculated by multiplying the prey proportion for each orca group by their biomass proportion of each orca group (Table 72). The single killer whale group was finally re-split into two distinct groups: transient orcas and resident orcas Table 72.  Estimated diet compositions of orca categories in PWS Prey Transient orcasa Resident orcasa Aggregated  orcasb Harbor seal 39.7 -- 5.7 Dall’s porpoise 46.7 -- 6.7 Harbor porpoise 13.7 -- 2.0 Salmon -- 75.0 64.2 Herring -- 25 21.4 a) Diets from Matkin and Hobbs (above); b) A generalized orca diet was derived by multiplying the diet proportions for each orca group by their biomass proportions in PWS (transient: 0.144; resident: 0.856). 60     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      The PWS Ecopath model working group initially decided to distinguish killer whales into two groups, transients and residents—as the diets of these two types of orcas are highly distinct in the wild (Ford 1994, 1999). Thus, two orca groups were included in the earliest versions of the PWS Ecopath model. However, these groups were aggre-gated to make room for other groups in the model that needed to be dis-aggregated, in order to avoid exceeding a maximum of 50 groups in the model. This made the task of orca dis-aggregation simple since original authored sections existed for these groups (see above) enhanced with additional prey information (McRoy and Wyllie Echeverria 1990). As the result of dis-aggregating orcas, the ‘small cetacean’ group (porpoises) and the ‘pinnipeds’ group needed adjustment in or-der to balance the model. The small cetacean P/B value was increased from 0.15 to 0.24, and the pinniped biomass was increased from 0.066 to 0.072 t⋅km-2. Sea otters were added to the diet of transient orcas at a low level (1.5%) because transient orcas have recently been observed to switch to consuming sea otters (Enhydra lutris) throughout the Aleutians and in Prince Wil-liam Sound (Estes et al. 1998, Hatfield et al. 1998, Garshelis and Johnson 1999). Otters were added to the transient orca diet at a small level because orcas could have a pro-found effect on otter populations in PWS, as they have in the Aleutians, even if otters are a small proportion of the transient orca diet (Estes et al. 1998). With this inclusion, tran-sient orcas could now switch to eating more otters when otter densities are high and other orca prey are scarce. Several other species of fishes were added to the orca diets at 1% levels fo facilitate prey switching (see veri-fication section).  Small Cetaceans Craig Matkin North Gulf Oceanic Society Homer, Alaska, USA Rod Hobbs Alaska Fisheries Science Centre NOAA, Seattle, Washington, USA The small cetacean group is composed primar-ily of Dall’s porpoise and harbor porpoise. Dall’s porpoise are most common during the summer and early fall and less common in late fall, winter and spring and may leave the Sound at that time (C. Matkin, unpublished data). Har-bor porpoise are more often sighted in the fall, winter and spring than the summer and often the largest groups appear in late March and early April (C. Matkin, unpublished data). For this exercise we assume that the temporal be-havior of these two porpoise species is com-plementary so that their peak populations are in September for Dall’s porpoise and March for harbor porpoise. The diet preferences shift to-ward those of Dall’s porpoise in the summer and fall and harbor porpoise in the winter and spring. Dahlheim et al. (unpub. data) estimate the density of harbor porpoise at 0.048 ⋅km-2 (CI = 0.030 - 0.066) in the region that includes Prince William Sound. Hobbs and Lerczak (1993) estimate the density of Dall’s porpoise in the Gulf of Alaska at 0.11 ⋅km-2 (CI = 0.07 - 0.16). An average Dall’s porpoise weighs 136 kg, and an average harbor porpoise weighs 55 kg (Wynne 1992). Thus, the annual biomass density estimate for small cetaceans is the mean of the biomass density estimates of Dall’s and harbor porpoise.  Harbor porpoise in captivity typically eat be-tween 4% and 9.5% of their body weight per day (Kastelein et al. 1997). We use 8% as an intermediate value for active porpoise and apply it to Dall’s porpoise also. Typically Dall’s por-poise feed on epi- and meso-pelagic squids and fishes (mostly <30 cm in length). In the north-western North Pacific Ocean Dall’s porpoise feed primarily on squids (Gonatidae) and lanternfish (Myctophidae) (Jefferson 1988). Although harbor porpoise feed primarily on gadoid and clupeoid fishes in the range of 10 - ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     61 25 cm, they also feed on a variety of cepha-lopods and other small fishes (Leatherwood et al. 1982, Kastelein et al. 1997). Relative fractions of components of the diets are un-known, so the above groups were roughly apportioned for this exercise based on avail-ability in the model. The general parameters used for this group are presented in Table 65.  DETRITUS Thomas A. Okey UBC Fisheries Centre, Vancouver, BC  For the purposes of this model, detritus is defined as organic carbon that is readily ac-cessible and usable by organisms in the sys-tem. The overall estimate for PWS detritus (134 tC⋅km-2) is the sum of a water column estimate (14 tC⋅km-2) and a conservative sediment estimate (120 tC⋅km-2)(see below). Detritus pools in the sediment and the water column were each split into nearshore and offshore detritus groups according to respec-tive sizes of these zones, thereby expressing each of the pools on a PWS-wide basis. The resulting split estimates were then combined within zones to estimate the detritus masses of nearshore (19.52 tC⋅km-2) and offshore (114.48 tC⋅km-2) detritus groups. Ap-proaches to calculating the detritus mass for these two groups are described in the fol-lowing sub-sections.  A third detritus group, ‘nekton falls,’ con-sists of ‘dead discards’ from PWS fisheries and salmon carcasses that re-enter the sys-tem. The ‘nekton falls’ group is discussed in the ‘fisheries’ section and the ‘adult salmon’ section in this volume.   Benthic detritus pool Feder and Jewett (1988) developed a carbon budget for Port Valdez, in PWS, in which the pool of benthic detritus (organic carbon; OC) was estimated to be 120 gC⋅m-2 (120 tC⋅km-2), based on an estimate by Naidu and Klein (1988). This is probably a minimum estimate for the larger Prince William Sound since the percent organic carbon measured in Port Valdez sediment was lower than almost all 24 PWS stations investigated in 1990 by Feder and Blanchard (1998) (Table 74), and because the organic carbon-limited benthos of Port Val-dez is less abundant than the benthos of the outer Prince William Sound and the adjacent shelf (Feder and Jewett 1988, H. Feder, UAF IMS, pers. comm., November 1999). Although the OC in Port Valdez sediment may be locally supplemented by some terrigenous sources (which are less labile, i.e., useful as food sources), seasonally-dense zooplankton densi-ties may disproportionately limit sediment OC thereby uncoupling planktonic production from the benthos. Benthic systems are more tightly coupled with overlying planktonic systems in certain nearshore nearshore continental shelf systems of Alaska, thereby receiving higher depositions of usable organic carbon (Greb-meier and Barry 1991), and this may be more true for outer PWS than Port Valdez (H. Feder, pers. comm., November 1999).   Estimates of degradable organic carbon in the mixed surface layer of sediment (0-20 cm) throughout the worlds oceans also indicate that the PWS benthic detritus estimate of 120 tC⋅km-2 may underestimate the pool of organic carbon in PWS; degradable organic carbon on conti-nental margins is estimated to range from 450-760 tC⋅km-2, while on abyssal plains it may range from 53-103 tC⋅km-2 (Emerson et al. 1987). However, the long term persistence of 'degradable' organic carbon detected in these sediments indicates that some fractions of the degradable OC in this surface mixed layer might be unusable by benthic fauna. In this light, benthic assemblages could exist in a state of carbon limitation even in situations of appar-ently excess organic carbon. Such limited ac-cessibility of organic carbon by benthic fauna is supported by evidence that only a low propor-tion of sea floor OC is used, and a high propor-tion is refractory (O'Reilly 1985). Therefore, the low proportion of sediment OC that is usable by benthos may offset the underestimate of organic carbon for PWS, due to the tendency for under-representation of PWS OC by Port Valdez data and the expression of OC in dry weight. Estima-tion of the extent of such compensation was not 62     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      attempted, due to uncertainty in these vari-ables.  Pelagic detritus pool The standing mass of pelagic detritus in Prince William Sound was estimated to be 14 tC⋅km-2 based on a contributed value for primary production of 228 tC⋅km-2⋅year-1 (doubled from P. McRoy’s contributed value of 114 tC⋅km-2⋅year-1, see Phytoplankton section) and a mean PWS euphotic depth of 25 (D. Eslinger, pers. comm., 5/1998) using the following empirically derived equation from Pauly et al. (1993);  log10D = -2.41 + 0.954 log10PP + 0.863 log10E     where D is the mass of standing pelagic de-tritus in tC⋅km-2, PP is the primary produc-tion in tC⋅km-2⋅year-1, and E is the euphotic depth in m.  This new estimate of the standing mass of pelagic detritus in PWS (14 tC⋅km-2) was used as wet weight in the PWS model, as an applicable conversion from C to wet weight was not identified for detritus. The model may thus underestimate the wet weight of standing detritus, but these values can be eas-ily modified in the future by users who can identify a useful conversion factor.   The pelagic detritus pool and the benthic pe-lagic pools were apportioned into nearshore and offshore detritus categories by calculat-ing the relative space for each pool in the area corresponding with each category. For example, 0.64% of the pelagic detritus space (volumetric) occurs inshore of the 20m iso-bath, while 99.36% occurs offshore (inte-grated to 300m mean depth). Likewise, 16% of the benthic detritus space (area) occurs inshore of the 20m isobath, while 84% occurs offshore. The splits of the pools apportioned to each zone were then combined for zone-specific estimates of detritus mass (Table 73). The large discrepancy between inshore and offshore values arises because the detri-tus pool in each area must be expressed on a sound-wide basis like the other groups in the model.    Input of terrestrial organic material An estimate of the flux of labile terrestrial or-ganic carbon to PWS sediments (9.6 t⋅km-2⋅year-1) was explicitly specified in the model. This import was split evenly between the inshore and offshore detritus groups (4.8 t⋅km-2⋅year-1 each) even though the area corresponding to the near-shore detritus group is 16% of the total PWS area, with the effect of the nearshore zone re-ceiving five times the input of terrestrial detritus as the offshore zone.   Several studies indicate that terrigenous sources of organic carbon make up a considerable pro-portion of the total organic carbon that reaches Table 73. Estimated mass of detritus pools in PWS split into nearshore and offshore zones. Zones are delineated by the 20m isobath. Density values for each area are expressed on a sound-wide basis for compatibility with other groups in the model. Mass of detritus (tC⋅km-2) Detritus group Pelagic Benthic Combined Nearshore 0.09 19.43 19.52 Offshore 13.91 100.57 114.48 Table 74. Percent organic carbon in sediment from locations in and around Prince William Sound and from earth's continental margins and abyssal plains. Means are presented with standard errors, except for range data from Port Valdez. Locations of samples Percent OC 13 PWS sites (40m depth)a 1.57 ± 0.44 13 PWS sites (100m depth) a 1.24 ± 0.28 26 PWS sites (40m & 100m) a 1.41 ± 0.26 Port Valdezb 0.1 –  0.6 4 GoA shelf and slope sitesc 0.55 ± 0.09 Earth's continental marginsd 1.02 Earth's abyssal plainsd 0.34 a) 1990 data from Feder and Blanchard (1998); b) range from Feder and Jewett (1988); c) from Seminov (1965) in Feder and Jewett (1986); these values likely underestimate %OC in the more OC-rich northern gulf region, which includes PWS; d) estimates from Emerson et al. (1987). ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     63 the sea floor of continental margins (e.g., Fahl and Stein 1997, Hedges et al. 1997, Macdonald et al. 1998). On the Canadian Beaufort shelf, terrestrial POC comprised 39% of the total POC reaching the sea floor (primary produced POC made up the re-maining 61%) (Macdonald et al. 1998), but in portions of lower Cook Inlet, which is adjacent to PWS, Lees and Driskell (1981) found that the majority of detritus on the bottom is from terrigenous sources (also see Feder and Jewett 1986). Thus, the 24 t⋅km-2⋅year-1 estimate of net flux of terrigenous organic carbon to the PWS sea floor is the product of Naidu's (1988) estimate of 48 t⋅km-2⋅year-1 for net flux of organic carbon to the sediments of Port Valdez and 50%, which represents the proportion of terri-genous origin OC in PWS sediments, based on the information presented above. How-ever, a large proportion of terrigenous or-ganic carbon (e.g., 60%) has been found to go unused (Macdonald et al. 1998 also see Hedges et al. 1997). The proportion of used to unused terrestrial OC might vary consid-erably among nearshore benthic systems, but its application here results in an adjusted estimate of 9.6 t⋅km-2⋅year-1 for the flux of liable terrigenous organic carbon to the PWS sea floor.  Adjusting assimilation efficiencies Unassimilated food / consumption ratios for plankton and benthic groups were changed from the 0.2 default to 0.4, meaning that the assimilation efficiencies (unassimilated food / consumption) were set to 60% rather than 80% for these groups. Default assimilation efficiencies of 80% (unassimilated food / consumption = 0.2) in the Ecopath software are unrealistic for de-tritivores and herbivores because much of the organic material (‘food’) consumed by these groups is of low quality in the sense that it has a low energy to mass ratio, or the energy can be difficult to utilize because some material is difficult to digest. Thus, assimilation efficiency takes food quality into account, and the efficiency of the feed-ing and digestion of the prey organisms. Lower trophic level organisms (i.e., especially herbi-vores) consume food that is of lower quality as defined here.  Erroneously high assimilation efficiencies caused early versions of the PWS model to be unrealistically 'tight' in energy terms. This ‘tightness’ was manifest in the detritus. Detritus in the system did not build up at all when as-similation efficiencies were too high. This was not acceptable because the system’s microbial community is implicit in the detritus, rather than explicit like all other biotic groups. In this model scenario, then, a considerable flux of ‘extra’ detritus must exist in order to feed the microbial community and account for other non-biological losses of OC.  This energetic tightness of the system exacer-bated the dynamic instability identified by Powell and Pimm (1999). Buildup of detritus as the result of the downward adjustment of as-similation efficiencies can be inferred from the decreases in the ecotrophic efficiencies of detri-tus groups (Figure 1) and considerable increases in flows to trophic level two (Figure 2), as con-secutive adjustments are made to plankton and benthic groupings. This surplus detritus should help to increase the stability of the model. However, the ‘instabilities’ experienced by Powell and Pimm (1990) have other causes in addition to the ‘energetic tightness.’ Notably, these analyses all assume strictly top-down sce-narios, i.e., assuming that all prey are always accessible to their predators, which is probably unrealistic. Such approaches, which feature full access of predators to prey, lead to artificial competitive exclusion, artificial cyclic behav-iors, and the loss of functional groups due to diet overlaps and due to the assumption that increases in mortality rates are proportional to increases in predation (C. Walters, UBC Fisher-ies Centre, personal communication, 18 Sep-tember 1999). This unrealistic behavior can eas-ily be generated on Ecosim (by setting the vul-nerabilities to ‘1’). However, our Ecosim runs were performed with lower vulnerability, corre-sponding to a mix of top-down and bottom-up interactions structuring the food web. Powell and Pimm (1999) also found that higher 64     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      trophic level organisms had higher instabili-ties than those at lower levels. This is to be expected in a fully top-down control model, because the complex stabilizing behaviors of higher trophic level organisms, such as com-pensatory prey switching and searching, are not accounted for. This is discussed by Powell and Pimm (1999; page 39, bullets 2 and 4). In this light, their finding that higher trophic levels in the early versions of the PWS model were less stable than lower tro-phic levels verifies reality, and indicates the need for inclusion of prey switching, refu-gia, spatial and temporal heterogeneity, and inefficiency of predation into dynamic simu-lations, as Ecosim and Ecospace do. Although the adjustment of assimilation ef-ficiencies led to a considerable increase of the detritus flux available to the microbial community, it did not change the character of the dynamics of particular working simu-lations in Ecosim. Further refinement of as-similation efficiencies in the future will fur-ther increase the usefulness of the Prince William Sound model.  PWS FISHERIES PWS fishery catches were explicitly as-signed to three fishery sectors: subsistence, recreational, and commercial. The commer-cial sector could be further broken down further into gear types, but this was not done at this stage.  Economic information about the fisheries can be incorporated in order to conduct fish-eries-related economic analyses, but no fish-eries economic information has been incor-porated into the model at this time.    Fishery Landings Estimates for PWS, 1994-1996 Commercial landings in PWS Estimates of commercial landing and dis-card rates in PWS are shown in Table 76. These estimates were developed from esti-mates of commercial groundfish (and shark) landings in Prince William Sound for 1994-1996 (Table 43; data provided by B. Bechtol, Alaska Department of Fish and Game); mean salmon landings in PWS from 1994-1996 (commercial and subsistence; Table 36; based on Morstad et al. 1997); and PWS herring land-ings (Figure 4; data provided by J. Wilcock, Alaska Department of Fish and Game); other information in this report (e.g. K. Frost, this volume), and the methods described in this sec-tion.  Recreational landings in PWS Scott Meyer Alaska Dept. of Fish and Game, Homer, Alaska Recreational landings estimates were compiled using two data sources: (1) the number of fish harvested was estimated through a postal survey using a large random sample of resident and non-resident license buyers (Howe et al 1995-97); (2) the species composition of the sport harvest and average weights were estimated through a port sampling program located in the Valdez harbor, conducted from late May - early September. Recreational groundfish catch esti-mates for the years 1994-1996 in PWS are shown in Table 75. Halibut average weights are based on lengths and a length-weight relationship generated by staff of the International Pacific Halibut Com-mission. Rockfish average weights were esti-mated from length measurements using species-specific or assemblage-specific length-weight relationships from 1991-1995 data from all over South-central Alaska. Lingcod average weights are based on measured lengths and a length-weight relationship using fish weighed in 1992-1996. One potential problem with rockfish estimates is that the species composition at Valdez may not be representative of PWS as a whole. An-other is that the species composition is very dif-ferent by user group (guided/unguided) in Val-dez, and that the proportions of harvest by each user group in Valdez are not representative of PWS as a whole. Guided anglers on charter boats in Valdez tend to fish the outer waters of ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     65 PWS around Montague and Hinchinbrook Islands, and catch mostly black and yel-loweye rockfish. Unguided anglers tend to fish inner, more protected waters, and catch more coppers and quillbacks. Estimates were not stratified by user group - when originally done no user group information was available for PWS harvest estimates. We explored the 1996 estimates to see the effect of this error; if 75% of the PWS catch is by unguided anglers (estimate from postal survey) and if we assume that the species composition of the rockfish harvest by pri-vate anglers all over PWS is similar to pri-vate anglers from Valdez, then we could be overestimating harvest biomass by 20% or more. This is because we are overestimating the harvest of larger black and yelloweye rockfish, and underestimating the harvest of smaller copper and quillback rockfish.  A third problem with the sport harvest data for Prince William Sound is that there is an unknown amount of error due to mis-reporting of areas fished, which would likely result in a slight underestimation of the hali-but, rockfish, and lingcod harvested from PWS waters. The error arises from the fact that some anglers report harvest by the port of landing rather than from the waters they fished.  Fishery Discard Estimates for PWS, 1994-1996 Thomas A. Okey UBC Fisheries Centre, Vancouver BC   Scott Meyer Alaska Department of Fish and Game, Homer, Alaska  Fishery discards that were as-sumed to be dead were ex-plicitly included and treated in the model. Spe-cies-specific estimates of discards are shown in Table 76 and were explic-itly incorporated into the discard input interface and added to species-specific landings estimates to obtain discard-adjusted estimates of catch (i.e., catch = landings + discards). Discarded marine organisms entered a new and separate detritus category, ‘nekton falls.’ This general name was used instead of ‘dead discards’ be-cause post-spawner salmon carcasses were also added to this group.  This detritus group is then fed on by detritivores, scavengers, and other fish predators, in which discards were specified in diet compositions.    Commercial discards. -- Discard flows from the commercial fishing fleet were estimated using reported discard information from the Alaska Department of Fish and Game (Charlie Trowbridge, unpublished data), but primarily from the fishery observer program of the Na-tional Marine Fisheries Service (M. Furuness; unpublished data; 10/1999). Annual discards for the PWS groundfish fisheries during the 1994-1996 period were estimated by calculating the ratios of discards in each species category to total catch (calculated from observed PWS cruises from 1994-1996 including hook and line, pot, and trawl fisheries) and applying these group-specific ratios to the total 1994-1996 PWS commercial catch estimate (provided by B. Bechtol, ADF&G; 8/1998), not including salmon and herring fishery catches. Discards associated with the herring fishery were then calculated based on rough estimates of discards for each PWS herring fishery sector (by J. Wil-cox; pers. comm., 25 Oct 1999). Salmon fisher-ies were assumed to have zero discards.   Recreational discards. – Estimates of the flow of discards from the recreational fisheries in PWS were developed based on the statewide postal survey of recreational fishing in which Table 75.  Estimated recreational groundfish landings in PWS (t round weight): Year Halibut Pelagic rockfish Demersal rockfish Slope rockfish Total rockfish Lingcod b 1994 338a 14.810 b 20.984 b 0.54 b 36.333 b 17.360 1995 383b 14.054 b 25.014 b 0 b 39.068 b 26.885 1996 429b 8.708 b 21.628 b 0.105 b 30.441 b 17.631 Mean 383 12.524 22.542 0.215 35.281 20.625 a) From Meyer (1995); b) Estimates from S.C. Meyer (Alaska Dept. of Fish and Game, pers. comm.). 66     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      respondents report the number of individuals of each species category that that they kept and released (Howe et al. 1995-1997). Mean weights of fish from the Valdez area were then applied to these PWS-wide data for to estimate total mass of ‘kept’ and ‘released’ fish. Ballpark estimates of mortality rates were then applied to the estimates of re-leased fish mass for a rough estimate of the mass of ‘dead discards,’ which enters the ‘nekton falls’ detritus category in the PWS model. The mortality rates that were applied to the released fish was 0.75 for the entire recreational rockfish complex and 0.05 for both lingcod and Pacific halibut.   A caveat on these estimates of landed and re-leased fish is required here because of the fol-lowing limitations: 1. Recreational fishers are not particularly skilled at marine fish identification; 2. Landings and reporting from Seward may not represent fishing activity in PWS; 3. Average fish weights calculated from the Valdez area may not accurately represent all of PWS.  The fate of these discards was assigned to the detritus group ‘nekton falls,’ as described pre-viously. The amount of post-run salmon car-casses re-entering the defined PWS system from adjacent rivers and spawning beds was taken as 25% of the estimated amount of es-caped adult salmon (56,174 t, or 6.201 t⋅km-2⋅year-1), resulting in 1.55 t⋅km-2⋅year-1 being imported into the ‘nekton falls’ detritus group. Table 76. Estimates of mean annual PWS fishery landings and discards (1994-1996).  Landings (t⋅km-2⋅year-1) Dead discards (t⋅km-2⋅year-1) Group Commerciala Recreationalb Subsistence Commercial Recreationalg Adult salmon 5.3726 - 0.0002c - - Adult Pacific herring 2.5455 - - 0.0551e - Pollock 1+ 0.1710 - - 0.0103f - Deep epibenthos 0.1430 - - - - Shallow demersals 0.0700 - - 0.0001 f - Pacific cod 0.0656 - - 0.0016 f - Pacific halibut 0.0268 0.0423 - - 0.0015 Sablefish 0.0184 - - 0.0008 f - Rockfish 0.0060 0.0039 - 0.0005 f 0.0020 Shallow large infauna 0.0030 - - - - Adult arrowtooth flounder 0.0004 - - 0.0019 f - Deep demersal fishes 0.0003 - - 0.0023 f - Sharks 0.0003 - - 0.0022 f 0.0019h Lingcod 0.0003 0.0023 - - 0.0002 Juv. arrowtooth flounder 0.0001 - - 0.0019 f - Juv. Pacific herring - - - 0.1189 e - Pinnipeds - - 0.0020d - - Totals 8.4234 0.0485 0.0022 0.1958 0.0037 a) adapted from data from Bill Bechtol (ADF&G; see Table 44) ; b) from estimates from S. Meyer (see Table 58 in this volume); c) from Morestad et al. (1997); d) from K. Frost (this volume); e) based on ballpark estimates by J. Wilcox, ADF&G (pers. comm., 25 Oct 1999); f) based on postal surveys by Howe et al. 1995-1997; g) group-specific discard ratios calculated from data provided by M. Furuness, NMFS; these ratios were applied to PWS landings data provided by B. Bechtol, AKF&G. h) see Hulbert (this volume) ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     67  Constructing and Balancing the PWS Model Thomas A. Okey Fisheries Centre, UBC, Canada The input parameters estimated in the previ-ous sections by the PWS Ecopath working group were entered into the ‘basic input’ and ‘diet composition’ spreadsheets of the Eco-path with Ecosim software application (available free from http://www.ecopath.org/). Fishery and mi-gration information provided in the previous sections, as well as detritus fate information was likewise entered into appropriate spreadsheets in the Ecopath software. In the PWS model, detritus fate information con-sists of the allocation of unconsumed and un-exported production from each group between nearshore (< 20 m depth) and off-shore detritus categories.   The PWS model contains 48 ecosystem components - presently 50 is the maximum number that the Ecopath software can ac-commodate. Groups were considered for aggregation when this limit had been reached and more groups needed to be added to achieve optimal realism. Groups that were most similar affinities, in terms of their basic inputs, were identified using an automatic aggregation routine, and then ag-gregated. Resident orcas and transient orcas were the first groups to be aggregated, as their trophic levels were similar.  After the initial model construction, the eco-trophic efficiency (EE) terms were exam-ined to evaluate the balance among compo-nents, and within the whole system. If a par-ticular group was ‘unbalanced’ within the model (i.e., when its ecotrophic efficiency was greater than 1), this indicated that bio-mass or production/biomass values for the group were underestimated, or that con-sumption by other groups was overesti-mated.  Estimates of these parameters for the con-nected groups can then be adjusted to bring the groups and the model into balance. However, since there are multiple connections among groups, a change in the estimate for a predator, for example, may in turn change the degree of balance with its predators in addition to its prey. Thus, a haphazard approach to model balancing may result in arbitrary parame-ter adjustments and lead to unnecessary erosion of model realism. A semi-systematic method was employed to address this problem by de-veloping a hierarchy of parameter adjustments; groups were ranked by degree of imbalance indicated by the amount that its EE exceeded 1.0; groups were also qualitatively ranked based on the balancer’s degree of confidence in the contributed parameters. The balancer’s degree of confidence was based on the weight of the available evidence.  One contributor pointed out that the general technique used to balance the first model itera-tion tended to adjust the biomass and produc-tion estimates upward in unbalanced groups rather than adjusting predator consumption rates downward (T. Dean, personal communication). This occurred for two reasons: (1) the assump-tion that biomass uncertainty is more likely un-derestimation than overestimation, (2) lowering consumption rates spreads effects of the ad-justment across all prey rather than just the un-balanced one, and (3) adjusting diet composi-tions also influences more than just the target groups and can erode model realism, as well as refine it. The potential interjection of bias relat-ing to these assumptions was considered during the final re-balancing. This resulted in improved agreement of the calculated phytoplankton pro-duction estimate with the initial phytoplankton production estimate provided by C. P. McRoy (see Phytoplankton section). Adjustments to the contributed parameter val-ues, made during balancing, are documented in the following sections. Phytoplankton and primary production Phytoplankton biomass was calculated by the model using P/B values of 190 and ecotrophic efficiency (EE) values of 0.95 for each of the two phytoplankton groups - nearshore and off-shore phytoplankton. The annual phytoplankton production required for the model to run is cal-culated by the model as 3,040 t⋅km-2⋅year-1. This 68     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      value is 38% greater than the initial phyto-plankton production estimate of 2,210 t⋅km-2⋅year-1 for PWS (see Phytoplankton section in this report). This initial estimate was based on samples from Port Valdez and Valdez Arm, and likely contained an under-estimating bias due to turbidity caused by glacial silts. The alternative explanation is that the current model is inflated (see above). Both could be true. Total produc-tion, including macroalgae and eelgrass, was calculated to be 3,666 t⋅km-2⋅year-1, again 35% greater than the initial 2,711 t⋅km-2⋅year-1 estimate for the total primary pro-duction of PWS from 1994-1996. Macroalgae and eelgrass The P/B for macroalgae and eelgrass was increased from 4 year-1 to 5 year-1 to balance the offshore detritus group. The biomass of offshore detritus was too low to support the demand on it by detritivores, and increasing the production of macroalgae/eelgrass effec-tively increases the contribution of drift al-gae to detritus.  Capelin This group was balanced by shifting preda-tion pressure from this group to juvenile her-ring, of which there was plenty in the model. Thus, predators were fed juvenile herring rather than so much capelin. In addition, the biomass value (0.231 t⋅km-2) was increased 60% to 0.367 t⋅km-2, and the P/B value was bumped from 3 to 3.5 year-1. Sandlance Again, predators of sandlance were fed lar-ger proportions of juvenile herring, and smaller proportions of sandlance. Herring The Adult herring group was balanced by increasing the biomass estimate by 10% to 2.810 t⋅km-2.  Squid The biomass estimate for squid (0.019 t⋅km-2) was highly uncertain mainly because the data used for the estimate came from a sam-pling program that was designed for estimating pollock (see squid section in this report). The biomass estimate was increased to 3 t⋅km-2 to balance this group, implying either that this group is undersampled, or that the contribution of squid to its specified predators is overesti-mated. Pollock The pollock age 0 group was balanced by in-creasing the P/B estimate to the upper end of its confidence range (see Table 31) and doubling the biomass from the upper end of its confi-dence range (0.05 t⋅km-2) to 0.11 t⋅km-2. The pollock age 1+ group was balanced by shifting some of the predation by its predators to other groups in the diet matrix. The aggregated bio-mass estimate (2.99 t⋅km-2) for pollock age 1+ was then multiplied by 2.5 to balance the group.  Sablefish The sablefish biomass estimate was multiplied by 1.5 (from 0.195 t⋅km-2 to 0.293 t⋅km-2) to balance this group. This original was thought to be an underestimate (see Sablefish section in this report). In addition, some of the contribu-tion of sablefish to the diets of predators was shifted to Pacific cod (which was less than fully exploited trophically), in order to balance the Sablefish group. Pacific cod The biomass estimate of 0.555 t⋅km-2 derived for Pacific cod in PWS resulted in a calculated ecotrophic efficiency (EE) value of less than 0.5 indicating that over half the PWS biomass of Pacific cod die of old age and become detritus rather than being preyed upon. The estimate adapted from the 1989 post-spill PWS multi-species trawl survey data (0.225 t⋅km-2) pro-duced an EE that was slightly over 1. Thus, a value intermediate of the two independent esti-mates (0.3 t⋅km-2) was used to achieve a reason-able EE  of 0.884.  Juvenile Arrowtooth Flounder  The given biomass value (0.57 t⋅km-2) was mul-tiplied by 1.5 to 0.855 t⋅km-2, a value well within the given confidence range. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     69 Deep Epibenthos This group was balanced by increasing the initial biomass estimate of 1.5 t⋅km-2 to 30 t⋅km-2 and increasing the P/B from 2 year-1 to 3 year-1 Rockfish Both the rockfish biomass estimate and pre-dation on this group are highly uncertain. This group remained unbalanced after the biomass estimate was multiplied by four, and predation on this group was shifted to other groups. If these adjusted input parame-ters do not underestimate biomass and P/B, and the group remains unbalanced (EE is greater than 1), then the model indicates that PWS rockfish are declining. An import value of 0.14 t⋅km-2⋅year-1 was entered in the migration sheet to represent rockfish de-cline, and allow model balancing (see Rock-fish section for further discussion).  Deep Demersals This group was balanced by increasing the biomass estimate by 10% to 0.96  t⋅km-2 and the P/B estimate by 30% to 1.008 ⋅year-1. Shallow Small Epifauna The biomass estimate used for this group was developed using data from the intertidal zone (Highsmith in Dean, this volume). This group was balanced by tripling the biomass from 8.7 to 26.1, based on evidence that shallow small epifauna can attain tremen-dously high abundances and biomass in the shallow subtidal (Vetter 1994 and 1995; Okey 1997). The P/B was also increased from 2.0 to 2.8 to obtain balance for the group. Macrofauna The deep small infauna group was balanced by doubling the biomass estimate and in-creasing the P/B from 0.96 ⋅year-1 to 3 ⋅year-1 The shallow small infauna group was bal-anced by multiplying its P/B by six, from 0.6 year-1 to 3.6 ⋅year-1. This value was then nudged 1 to 3.8 ⋅year-1 Deep large infauna This group was balanced by increasing the ini-tial biomass estimate (16.2 t⋅km-2) by 75% to 28.35 t⋅km-2. Shallow large infauna The ecotrophic efficiency (EE) for shallow large infauna was less than 0.25. To increase the utilization of this component to a more real-istic level, the ratio of deep and small large in-fauna (clams) in the sea otter diet was adjusted from equal allocations of 80% (40% and 40%) to 70% shallow prey (<20 m depth) and 10% deep prey (>20 m depth). This adjustment as-sumes that otters exploit shallow clam resources before venturing deeper.  Zooplankton The biomass estimate of offshore omnivorous zooplankton was increased by 60% from 15.4 t⋅km-2 to 24.64 t⋅km-2, and its P/B was increased to 11.06 ⋅year-1 (40% above the 7.9 t⋅km-2 value given for nearshore omnivorous zooplankton) to achieve balance in that group. Its Q/B was then increased from 17 to 22.13 year-1 to limit the P/Q to 0.5 year-1, but this is more in line with the estimates for nearshore omnivorous zoo-plankton. The P/B of offshore herbivorous zoo-plankton was increased 60% from 15 t⋅km-2 to 24 t⋅km-2. Nearshore omnivorous zooplankton biomass was increased 30% from 0.079 ⋅year-1 to 0.103 year-1 The biomass of nearshore her-bivorous zooplankton was increased 40% from 0.097 year-1 to 0.136 year-1  Small Cetaceans To balance this group relative to the specified predation by orcas, the maximum specified biomass range value (0.0132 t⋅km-2) was in-creased by 15%, and the P/B value was in-creased from 0.10 to 0.15 year-1.  Orcas Five new categories were added to the Orca diet at a symbolic 1% level because they were iden-tified by McRoy and Wyllie Echeverria (1990; see below) as part of the PWS Orca diet. This refinement was not necessary to balance the model, but it was preferred, as it allows prey switching in the dynamic simulation routines Ecosim and Ecospace, and presumably make 70     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      for more realistic simulations. Sea otters were added to the transient orca diet (see orca section).  Pinnipeds This group was balanced by using the upper end of the given density estimate for harbor seals in PWS, are 0.06 t⋅km-2 (Frost, this volume), and adding 10% to account for Steller sea lions, resulting in a biomass esti-mate of 0.066 t⋅km-2 for pinnipeds in PWS.  Lingcod   The lingcod group was balanced by multi-plying the initial biomass estimate by 3.5. Salmon fry 0-12 cm A P/B value of 7.154 year-1 was used instead of the calculated 9.844 year-1. It is 27% smaller.  Sharks The increased consumption by sharks in the model led to several groups consumed by sharks going unbalanced. This necessitated adjustment of the shark diet composition rather than adjusting parameters of its prey groups without justification. Adjustment of shark diet composition is justified due to the levels of uncertainty in its construction. In general, this diet balancing involved shifting predation pressure from the unbalanced group to the adult salmon group, which still had a relatively low EE value (0.454). This particular approach to balancing was strongly supported by August 1999 observa-tions of high incidences of adult salmon in the diets of sleeper sharks in addition to salmon sharks (L. Hulbert, pers. comm., 21 September 1999; n = 16 stomachs). Table 77 indicates pro-portion of shark diet composition shifted to adult salmon according to species.  Final input parameters The final input parameters and detritus fate in-formation for the PWS Ecopath model, after the above balancing adjustments were made, are listed in Table 78; input diet composition  val-ues are listed in Appendix 5, and three immigra-tion terms were entered: Adult salmon: 3.0 t⋅km-2⋅year-1; Eulachon: 3.0 t⋅km-2⋅year-1; and Rockfish: 1.4 t⋅km-2⋅year-1 (the value for rock-fish represents decline in the group and was needed for balancing).  Table 77. Proportion of shark diet composition shifted from the fol-lowing species to adult salmon. Species Proportion Sharks 0.015 Pacific halibut 0.020 Sablefish 0.040 Adult arrowtooth 0.020 Juv. arrowtooth 0.008 Pacific cod 0.060 Deep demersals 0.004 Rockfish 0.006 Total 0.173 ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     71 Table 78. Basic input parameters and detritus fate for the Prince William Sound model, 1994-1996. TL is the trophic level calcu-lated by Ecopath, OI is the omnivory index indicating the degree of omnivory, P/B is production/biomass, Q/B is consump-tion/biomass, and EE is the ecotrophic efficiency expressing the proportion of the production lost to export or predation. Detritus fate is a percentage allocation of the remaining production between detritus pools. Values in bold were calculated by Ecopath; val-ues not bolded are empirically-based input estimates, contributed by a collaboration of experts on PWS (Okey and Pauly 1999). Detritus fate (%) Group  Trophic level  OI  Biomass (t⋅km-2) P/B (year-1) Q/B (year-1)  EE  N. falls  Inshore  Offshore   Export  Transient Orca  5.4 0.01 0.001 0.05 6.04 - 1 - 50 49  Resident Orca  4.9 0.21 0.015 0.05 8.67 - 1 - 50 49  Sharks  4.5 0.98 0.662 0.10 7.00 0.753 1 - 99 -  Halibut  4.5 0.36 0.677 0.32 1.73 0.865 1 - 99 -  Porpoise  4.5 0.20 0.015 0.24 29.20 0.989 1 30 69 -  Pinnipeds  4.4 0.14 0.072 0.06 25.55 0.994 1 30 69 -  Lingcod  4.3 0.35 0.077 0.58 3.30 0.816 - 40 60 -  Sablefish  4.0 0.87 0.293 0.57 6.42 0.774 - - 100 -  Adult flounder  4.2 0.12 4.000 0.22 3.03 0.792 - - 100 -  Adult salmon  4.2 0.06 1.034 6.48 13.00 0.660 - 30 - 70  Pacific cod  4.1 0.47 0.300 1.20 4.00 0.936 - - 100 -  Juv flounder  4.0 0.12 0.855 0.22 3.03 0.956 - - 100 -  Avian raptors  3.9 1.58 0.002 5.00 36.50 - - 25 - 75  Seabirds  3.8 0.55 0.011 7.80 150.60 0.425 - 40 40 20  Deep demersals  3.8 0.80 0.960 0.93 3.21 0.984 - - 100 -  Pollock 1+  3.8 0.25 7.480 0.71 2.56 0.982 - - 100 -  Rockfish  3.7 0.26 1.016 0.17 3.44 0.969 - 20 80 -  Baleen whales  3.7 0.16 0.149 0.05 10.90 - 1 - 99 -  Juv. salmon  3.5 0.31 0.072 7.15 62.80 0.931 - 30 70 -  Nearshre demersal  3.3 0.24 4.200 1.00 4.24 0.710 - 100 - -  Squid  3.3 0.01 3.000 3.00 15.00 0.938 - - 100 -  Eulachon  3.2 0.63 0.371 5.00 18.00 0.998 - 40 20 40  Sea otters  3.2 0.18 0.045 0.13 117.00 0.005 - 50 50 -  Deep epibenthos  3.2 0.62 30.000 3.00 10.00 0.958 - - 100 -  Capelin  3.1 0.02 0.367 3.50 18.00 0.962 - 50 30 20  Adult herring  3.1 0.01 2.810 1.54 18.00 0.955 - - 100 -  Pollock 0  3.1 0.01 0.110 2.34 16.18 0.945 - 50 50 -  Shal large epibenth.  3.1 0.03 3.100 2.10 10.00 0.750 - 80 20 -  Sea ducks  3.1 0.00 0.005 0.20 450.50 - - 40 40 20  Sandlance  3.1 0.01 0.595 2.00 18.00 0.841 - 50 50 -  Juv. herring  3.0 0.01 13.406 0.73 18.00 0.919 - 30 70 -  Jellies  3.0 0.11 6.390 8.82 29.41 0.004 - 10 90 -  Deep sm infauna  2.3 0.23 49.400 3.00 23.00 0.916 - - 100 -  Near omni-zoo  2.3 0.19 0.103 7.90 26.33 0.980 - 70 30 -  Omni-zooplank  2.3 0.19 24.635 11.06 22.13 0.978 - 10 90 -  Shal sm infauna  2.2 0.18 51.500 3.80 23.00 0.941 - 100 - -  Meiofauna  2.1 0.11 4.475 4.50 22.50 0.950 - 20 80 -  Deep lg infauna  2.1 0.09 28.350 0.60 23.00 0.931 - - 100 -  Shal sm epibent  2.1 0.05 26.100 2.30 10.00 0.975 - 70 30 -  Shal lg infauna  2.0 0.00 12.500 0.60 23.00 0.516 - 100 - -  Near herbi-zoo  2.0 0.00 0.136 27.00 90.00 0.978 - 70 30 -  Herbi-zooplankt  2.0 0.00 30.000 24.00 50.00 0.976 - 10 90 -  Nearshr phytopl  1.0 0.00 5.326 190.00 - 0.950 - 70 30 -  Offshore phytopl  1.0 0.00 10.672 190.00 - 0.950 - 10 90 -  Macroalgae & grass 1.0 0.00 125.250 4.00 - 0.135 - 50 50 -  Nekton falls  1.0 0.21 2.000 - - 0.953 - 80 20 -  Nearshore detritus  1.0 0.30 19.520 - - 0.542 - - 100 -  Offshore detritus  1.0 0.46 114.480 - - 0.587 - - - 100 72     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996        Verification of web structure Thomas A. Okey Fisheries Centre, UBC, Canada The food web structure (diet compositions) of the PWS Ecopath model was compared with food web diagrams constructed for 35 individual species by McRoy and Wyllie Echeverria (1990). This approach was used to verify the existence of trophic links, not the magnitude of flows, though indications of ‘principal prey’ by these authors shed some light on relative flows. Inconsistencies are highlighted in the following sections; no comment was made when food webs were consistent. Seabirds Diet compositions of seabird species in Table 61 correspond well with individual food webs in McRoy and Wyllie Echeverria (1990), except for tufted puffins which the latter authors suggest consume capelin, sandlance, squid, euphausiids, and nearshore small epibenthos. Avian predators McRoy and Wyllie Echeverria (1990) in-cluded juvenile fishes, forage fishes, young sea otters, young seals, and young sea lions in the diet of bald eagles. These groups are represented by the ‘other marine’ category in Table 61, though predation on some of these groups is probably rare. Small cetaceans McRoy and Wyllie Echeverria (1990) indi-cate that salmon is a major prey of Dall’s porpoise, whereas salmon were not consid-ered part of the small cetacean diet in the current model. Beluga whales were also not included in the small cetacean category for Prince William Sound. It was decided that visitation of individuals from the Cook Inlet stock (~800 individuals) would have pro-duced a PWS biomass too low to exist as a group in the model (and likely too low to be functionally important). Pinnipeds Squid was specified as 6% of the harbor seal diet in the PWS Ecopath model, whereas it was not included by McRoy and Wyllie Echeverria (1990). Squid was, however, in-cluded in the diet of northern fur seals. Baleen whales (humpback whales) McRoy and Wyllie Echeverria (1990) in-clude both capelin and pollock as prey of humpback whales, whereas these two spe-cies were left out of the diet specification for humpback whales in the PWS Ecopath model. Conversely, sandlance was specified in the model’s humpback whale diet, yet it was not included in McRoy and Wyllie Echeverria (1990). Orcas McRoy and Wyllie Echeverria (1990) iden-tified five categories of primary prey of Orcas not specified in the PWS model: hali-but, sablefish, Pacific cod, pollock, and greenlings. The model specified one cate-gory not included in their compendium—herring. Trophic levels Another aspect of food web structure that was verified are the trophic level estimates for PWS generated by Ecopath. These were found to closely correspond to those esti-mated by the ratios of stable nitrogen iso-topes (Kline and Pauly 1998). ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                   73  73 Ecosim and Ecospace Methodology Once the Ecopath model was constructed and balanced, the effects of perturbations, or changes in fishing, were simulated over a selected time horizon (typically 10 years) by adjusting the mortality regime over that same period for one or more component of the ecosystem. This method was used to simulate a number of ‘what if’ scenarios provided during the model specification workshop in March 1998. These are listed in Box 2. Adjusting the mortality regime enables a simulation of responses from connected components of the ecosystem based on their relationships and the rates of trophic flow among them. These mortality adjustments were drawn into Ecosim’s graphical inter-face of mortality rate, using a mouse. Simu-lations can be repeatedly re-run as the user adjusts mortality rates. Scenarios were saved and archived within the software to be ac-cessed and re-run at any time in the future.   The recently-developed Ecospace routine (Walters 1998, Walters et al., in press) was used to simulate changes in spatial distribu-tions of Prince William Sound groups start-ing with information on habitat preferences as well as spatial distributions of habitats and organisms provided by contributors. Ecospace simulates dynamic, two-dimensional re-distribution of ecosystem components based on trophic interactions (flow) among organisms, their relative pref-erences for spatially-specified habitats, and their movement rates and vulnerability to predators in the various specified habitats. A spatial representation of Prince William Sound and its various marine habitats was created in the spatial ‘mapping’ interface of the Ecospace routine. This was done by overlaying a geographically-referenced grid system over a scanned map of Prince Wil-liam Sound using a computer drawing appli-cation. The grid was then reproduced with an Excel spreadsheet and hard copies were laid over one another. Boxes that covered mostly land were colored in and then trans-ferred to Excel’s coloring function. This map was then used for electronic and hard-copy distribution to contributors as a stan-dardized template for specifying the distri-butions of organisms and habitats, on a reso-lution useable in the Ecospace application. This same template was used to create the Box 2. Hypothetical ‘what if’ scenarios for simulations runs 1. What if fishing pressure on herring increases or decreases; what if there is one stock of herring? two? three? 2. What if somebody decides to fish sandlance or capelin? This is probably far-fetched, but model simulations would likely show important trophic impacts of removing important forage fishes. 3. What if an earthquake raises the upper 10m of intertidal above sea level? 4. What if PWSAC goes broke and the hatcheries close? 5. What if there is another oil spill? 6. What if human impacts from the road to Whittier result in damage to intertidal habi-tats in the western part of PWS? 7. What if recreational fishing pressure removes 90% of the rockfish from PWS? 8. What if there is a major warm-water episode for 2 years, such that the upper 200 m of water over the shelf in the GoA is elevated by 2 oC ?  9. What if the bloom and sustained productivity lasts only for 3 weeks instead of the usual 12 weeks in PWS ? 10. What if the harbor seals continue to decline at 8% per year ?  11. What if dungeness crab return to PWS? 12. What if salmon prices drop or increase? 13. What if pollock disappear from PWS? 14. What if salmon farming were allowed in PWS? 15. What if a road were established to Cordova? 16. What if cruise ship traffic increases into Cordova? 74     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      map of land and habitats in Ecospace.  An area 178 km wide (E-W) and 167 km tall (N-S), encompassing Prince William Sound, was re-constructed by coloring grid cells in the Ecospace mapping interface. Each grid cell was designated as land, or one of the following six habitats: 1. Nearshore rocky; 2. Shallow Soft Bottom; 3. Sandy/Muddy intertidal;  4. Open Water/Deep Soft Bottoms; 5. Productive entrances; 6. Deep fjords. This specification was quickly accomplished by drawing with a mouse while toggling different category tools. Preference for each habitat was then speci-fied for each ecosystem component in the model, by going to ‘ecology parameters’ in the menu, as were other parameters such as ‘relative movement in bad habitat’, and ‘vulnerability to predation in bad habitat’. Default settings proposed by the Ecospace software, and based on the developer’s ex-perience, were used as a guide for setting these parameters during this initial analysis. Redistribution of biomass densities of each species in PWS was then evaluated in the context of different hypothetical scenarios of system changes and forcing.  RESULTS Thomas A. Okey Fisheries Centre, UBC, Canada The goal of the construction of a balanced trophic model of Prince William Sound us-ing Ecopath was to incorporate all the biotic components of the marine ecosystem (im-plicitly or explicitly) in a cohesive descrip-tion of the food web and to provide a func-tional venue for synthesis of existing ecosys-tem information to achieve that goal. This collaborative effort resulted in the construc-tion of the most explicit Ecopath model con-structed to date, including a total of 48 eco-system components (Figure 6; also see Okey and Pauly 1999). Trophic connections have been omitted from the box diagram (Figure 6) because these connections are too numer-ous to display in a useful manner. The na-ture of the trophic interactions in this model is summarized by a matrix of mixed trophic impacts (Figure 7).  The Ecopath model of Prince William Sound was designed to view the system at a particular resolution, which is necessarily lower than that possible in concentrated studies of smaller sub-sections of the sys-tem, and thus some of the detailed informa-tion collected by other research projects in PWS cannot be incorporated into this analy-sis. However, the approach is designed to enable a whole-system view using parame-ters that are basic to understanding popula-tions and the ecosystem, and which should be information rich and highly refined. Rela-tively accurate estimates of these parameters were available, or calculable, for enough groups to enable construction of a useful PWS model at the intended resolution. It is important to highlight the potential for refinement and learning provided by a bal-anced trophic model. This model of PWS represents a possible scenario of relation-ships among groups (during the modeled period) as defined by the contributed infor-mation and the known constraints to the sys-tem. Estimates for relatively unknown com-ponents, for which confidence in the esti-mates is low, were refined based on these constraints and subsequent balancing.  How-ever, this is just a focal point in the refine-ment process, the most useful result of which is the re-visiting of ecosystem data and estimates by researchers and research program managers. In this sense, the bal-anced trophic model of PWS may become a powerful tool for formulating questions and guiding research in PWS and beyond. Thus, the Ecopath model of PWS, using the Eco-sim and Ecospace simulation routines, should be useful in several ways beyond the goal of synthesis of existing information into a comprehensible description of the sys-tem’s biotic components and their trophic ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                   75  75 relationships.  The Ecopath model of PWS presented here is a static model that represents the average annual state of the ecosystem based on the three years of data that were included in this synthesis (1994-1996). Ecosystem compo-nents fluctuate in the real PWS, as in other systems, and these fluctuations are driven by both biotic interactions and physical forcing. However, the Ecosim routine enables a simulation of environmental forcing, both at seasonal and longer term scales. For exam-ple, seasonal changes in primary production can be imposed on the system, while a 20-year shift in the regional climate regime can be simultaneously imposed. These physical forcing functions can be applied to selected groups in the model. Fluctuations in these forced groups would subsequently drive fluctuations in other components of the sys-tem.  The next step in the development of the PWS model is to incorporate explicit sea-sonality into the model using empirical sea-sonal data for various groups and interpo-lated data for groups without explicit sea-sonal data. Such interpolation can be under-taken by, for example, using empirically-derived relationships between temperature and metabolism (i.e., consumption rates). However, this can best be done as an exer-cise of its own, using the version of ecopath recently developed to accommodate explic-itly seasonal inputs (Martell 1999). Charac-terization of seasonal changes currently con-sist of estimates and discussions within indi-vidual sections of this report, and to analyses using Ecosim forcing functions on the cur-rent (annual average) Ecopath model of PWS.    76     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      76  Figure 6. Biotic components of the balanced trophic model of Prince William Sound, Alaska displayed on a trophic level scale (vertical axis). Biomass (B) is displayed for each component in t⋅km-2; production (P), consumption (Q), and total input (TI) are expressed in  t⋅km-2⋅year-1. Trophic flows are not displayed here, as there are too many connections for this format. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     77  Figure 7. Mixed trophic impacts of groups in the PWS ecosystem model, representing the direct and indirect impacts that a small increase of the groups on the vertical axis would have on those on the horizontal axis. The black bars above the lines are positive impacts (facilita-tion) whereas the shaded bars extending below the lines are negative impacts (inhibition). The routine that generated this graph was adapted from the input-output analysis of Leontief (1951).  78     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      One of the main uses of a balanced trophic model approach, like Ecopath, is the insight it can provide into indirect effects of known, or predicted, changes in certain parts of the system (Gaedke 1995, also see Wootton 1994 and Menge 1995), and thus, it is a way to gain understanding of the functioning of parts of the defined ecosystem, within the connected system. Both the description by Ecopath and the prediction by Ecosim and Ecospace can be useful in any part of the scientific process, from synthesis and sum-mary to research design and hypothesis for-mulation. An obvious direct application of this approach is resource management and planning.      Construction of the Ecopath model can also highlight groups in the ecosystem for which little information is known. These ‘weak links’ in the model limit our understanding of the system even though they get more refined through the modelling exercise. These can then be a focus of future research.   Temporal simulations of perturba-tions The Ecosim routine can be used to simulate perturbations, which could be stated as ‘what if’ scenarios, like as those provided during the collaborative model development and shown previously. The approach can be used to simulate of changes in the relative biomass trajectories of ecosystem compo-nents over a specified time horizon in re-sponse to specified changes in mortality rates for one or more components. These temporal simulations have been the feature of Ecosim, but additional simulation rou-tines have been added to the software. Of particular note is the spatially-explicit dy-namic simulation routine Ecospace.  Three caveats are useful here for under-standing the output of the Ecosim runs in this section: (1) simulated responses in eco-system components are the result of biotic interactions only, and do not include any responses of, or control by, physical forcing in the environment. Physical forcing (e.g., oceanographic regime shifts) can be simu-lated in Ecosim by various methods, but this was not necessarily the focus of the simula-tion examples herein; (2) Ecosim runs are based on one possible scenario for the PWS food web, albeit a likely scenario based on the information at hand from 1994-1996. Ecosim enables prey switching based on prey availability and other factors (see Wal-ters et al., in press), even though the starting point of each simulation is the generalized Ecopath model of the system in which diet compositions are specified; (3) the extent to which biological forcing and subsequent cascading effects, and system destabiliza-tion, in the simulations resembles the dy-namics of the real PWS depends on the de-gree of interactive plasticity as well as prey vulnerability, both of which can be adjusted and refined.   An unlimited variety of simulations can be conducted as researchers and other individu-als explore the model food web. Only a few simulations are provided below as examples of the different ways that the approach can be used. Example 1: Removing sharks This example illustrates how the ecological role of a single group can be explored. Fig-ure 8 illustrates the output of a dynamic simulation of removing sharks from Prince William Sound. The biomass of some spe-cies increase in response to shark removal, while the biomass of other species decrease, according to the trophic relationahips speci-fied in the Ecopath model. In this simula-tion, an important suite of predators (sharks) is removed from the system, and several other fish are predicted to increase in re-sponse. These are either competitors or prey of sharks, or both. The degree of importance of the types of relationships can then be ex-plored. In addition to the species that in-creased in response to shark decreases, sev-eral species declined. These include juvenile and adult herring, capelin, juvelile pollock, squid, rockfishes, other shallow and deep demersal fishes, and avian raptors. These are the types of organisms that might be facili-ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     79 tated by the presence of sharks in the PWS system that existed between 1994 and 1996). (Note: Caution is advised when interpreting the salmon trajectory since salmon are tran-sients in the system). Example 2: Changes in fishing effort The example depicted in Figure 9 illustrates how the Ecopath with Ecosim approach can be used to simulate the responses of biota to increases or decreases in the fishing mortal-ity associated with a particular fishery. This example illustrates changes in the biomass of several species of high-trophic level predators in response to increases or de-creases in fishing mortality. This simulation indicates that the fisheries of Prince William Sound not only directly influence the bio-mass of particular species (i.e., herring and salmon), but they also compete with other predators for food to the entent that this competition influences the biomass of these mamalian, avian, and fish predators. It was initially surprising that such strong indirect effects of the fishing in PWS was indicated by these simulations, given that the annual flow of biomass to fisheries is 0.24% of the overall flows of biomass in the system. Such simulations provide insights into the effects of diverting forage fish energy in these types of marine ecosystems. More detailed ap-proaches can be taken to pursue the phe-nomenon of 'trophic interception' revealed by this broad-scale simulation.  Example 3: Catastrophic disturbances Figure 10 depicts an analysis of broad-scale an more complex disturbances in which more than one biotic group are directly im-pacted and the indirect responses other or-ganisms can be explored in addition to the more general character of the response of the biotic system as a whole. In this case, the responses of the food web to three scenarios of catastrophic disturbances were compared. These consisted of  (a) the great Alaskan earthquake of 1964 (magnitude 9.2), which shook and tilted Prince William Sound caus-ing tsunamis, and which mostly impacted lower and mid trophic level organisms; (b) Scenario #1 of the Exxon Valdez oil spill, based on documented impacts of the spill, which focused on impacts to upper trophic level organisms, and some lower trophic level organisms of the intertidal; and (c) Scenario #2 of the Exxon Valdez oil spill, also based on documented impacts of the spill, but complemented with likely impacts of the spill that were not documented, in which both upper and lower trophic level organisms were impacted.  Good information on the direct effects of these complex physical disturbances is nec-essary for the simulations to be meaningful. This type of analysis is, thus, less straight-forward than explorations of ecosystem the roles of single species. This analysis is pre-sented here to illustrate the types of simula-tions that are possible. This simulation is, in effect, an exploration of the trophic charac-ter of disturbance. It indicates that ecosys-tems can recover rapidly to disturbances that affect mostly lower trophic levels; they re-cover more slowly when disturbances affect mostly upper trophic levels; and they may well stabilize at alternate stable states when disturbances impact a mix of high and low trophic levels. This analysis indicates that the working assumption that Prince William Sound is recovering from the Exxon Valdez Oil Spill may be a reckless assumption. These simulations indicate that Prince Wil-liam Sound, or other ecosystems, might not recover from disturbances that are severe enough across a broad range of trophic lev-els. Prince William Sound might 'stabilize' in an altered state.    80     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                  Figure 8. Simulated removal of sharks from Prince William Sound, revealing potential indirect trophic cascading effects of sharks. This figure shows increases and decreases in biomasses of various species (i.e., groups) in response to removal of sharks, based on their trophic relation-ships with sharks, or species affected by sharks (e.g., prey, competitors, etc.)    Figure 9. Simulation of changes in the biomass of Prince William Sound biota along a continuum of change in commercial fishing effort. Current commercial fishing effort corresponds with the '1' on the horizontal axis.  The simulation indicates that several predators would be more abundant if less fish were commercially caught; these same species would be less abundant at higher levels of fishing. Two species are predicted to go below 1/8 the current abundance at twice the current levels of fishing. 012340 1 2 3 4 5 6 7 8 9 10Time (years)Relative biomass Pacific codSalmonHalibutSablefish LingcodSharksSalmon Sharks Seals Resident orcaEagles Halibut HerringPorpoise Herring 1 0 3 1/8 1/3 1 3 8 Relative fishing effort Biomass / original biomass Salmon Resident orca 2 ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     81   Figure 10. Simulations of three catastrophic disturbances in Prince William Sound, Alaska: (a) the great Alaskan earth-quake of 1964 (magnitude 9.2), which shook and tilted Prince William Sound causing tsunamis, and which mostly im-pacted lower and mid trophic level organisms; (b) Scenario #1 of the Exxon Valdez oil spill, based on documented impacts of the spill, which focused on impacts to upper trophic level organisms, and some lower trophic level organisms of the in-tertidal; and (c) Scenario #2 of the Exxon Valdez oil spill, also based on documented impacts of the spill, but complemented with likely impacts of the spill that were not documented, and in which both upper and lower trophic level organisms were impacted. a) Alaskan earthquake of 1964 (mag. 9.2) c) Exxon  Valdez oil spill scenario #2 b) Exxon Valdez oil spill scenario #1Biomass / original biomass 3 3 3 1 1 1 0.3 0.3 0.3 0 5 10 15 20 25Time (years) 82     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Spatially explicit simulations The Ecopath with Ecosim software now in-cludes a routine that enables temporal simu-lation of biomass trajectories in a spatially-explicit context (Walters et al., in press). Biomasses of the various components in a food web redistribute themselves spatially according to the interface between habitat characteristics and trophic interactions. Fur-thermore, spatially-explicit questions or management options can be explored. A simulation using the Prince William sound model is presented here only for the purpose of providing an example of how the Eco-space routine can be used. The particular simulation presented here was not intended to be precise; it was intended as an example of the general types of responses that might be encountered during such simulations. Ecospace example : a marine protected area Ecospace was used with the Prince William sound model to explore the effects of a large marine protected area encompassing ap-proximately half of the 9,059 km2 sound. The first step in preparing an Ecospace sce-nario is to delineate the various types of habitats in the ecosystem. The numbers and differently colored (or shaded) cells in Fig-ure 11 correspond with the habitat types pre-sented above. This type of information was collected during early stages of this project by inference (in the form of spatial distribu-tion patterns of the organisms) and more directly (information about habitats). Step two involves assigning habitat preferences to each biotic component (species or guild). Preferences are defined in the form of rela-tive movement rates, relative vulnerability to predation, and relative feeding rates in the various habitat types.     Figure 11. Diagrammatic map of Prince William Sound. Land areas are shown in black. Specified habitats are numbered and color coded. A simulated marine protected area (MPA) is delineated in the Southwestern part of PWS by the dotted white lines.  ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     83 Figure 12. Simulated spatial re-distributions of three of the 48 groups in the PWS model based on trophic interactions in preferred versus non-preferred habitats; with unrestricted fishing effort (left hand column), and five years after the simulated establishment of  a ma-rine protected area (right hand column). Colors (or shadings; right panel) represent changes in biomass density, where reds are increases in biomass density, blues are decreases, and green corresponds with  the biomass density of the uniformly distributed, pre-simulation model. Before protected area 5 years after MPAHalibut LingcodSalmon 84     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Some of the results of this simulation of the effects of a marine protected area are shown in Figures 12 and 13. Three of the 48 groups in the PWS model are shown to have in-creased five years after the establishment of this large marine protected area. For this simulation, the biota in the model were given five years to equilibrate with each other in the specified habitats before a five year establishment of the marine protected area. Not shown in these figures is the re-sulting spatial distribution of fishing fleets, which began congregating along the boundaries of the MPA. Also not shown in this simulation are the responses of PWS rockfish (Sebastes). An additional simula-tion is needed for meaningful spatial analy-sis of this group of species in PWS (which is probably declining).  Ecopath and Resource Manage-ment Thomas A. Okey Fisheries Centre, UBC, Canada The Ecopath trophic modelling approach consists of both static and dynamic model-ling in a windows-based software package. The static Ecopath model is a quantitative description of the trophic flows in an eco-system averaged over a pre-defined area and time period. It includes all components of an ecosystem (aggregated into fifty ‘boxes’ or less), so it can be used as a focal point for collection of broad information about an ecosystem, including estimates of basic population, production, and consumption information, fisheries information, and other migration and trend information. The static Ecopath model is the foundation upon which effects on all components of ecosystems can Figure 13. Temporal biomass trajectories associated with spatial re-distribution of PWS biotic components both before and after the establishment of a simulated marine protected area. Most groups equilibrate in the specified spatial arena, and some groups increase after establishment of a marine protected area in which fishing is excluded.  0 5 10 1 0.1 10 MPA Time (years) Biomass / original biomass ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     85 be considered when analyzing the effects of human activities, whether in the past, pre-sent, or future. Ecosim and Ecospace are dynamic model-ling routines that use the information in Ecopath models to simulate ecosystem changes resulting from natural or anthropo-genic changes in the described system. For example, Ecosim is used to predict changes in the biomass of all the groups of organisms when the mortality of one or more group is changed (thereby changing the biomass of that group). These predicted changes are based on the trophic relationships described in the Ecopath model.  A typical Ecosim analysis consists of an in-vestigator increasing the mortality rate of one organism over a time horizon (say 10 or 20 years) by drawing (with a mouse) a mor-tality trajectory in the graphical interface for input. This increase in mortality might simu-late an increase in fishing, or a possible ef-fect of increased sediment runoff, effluent, or collecting and food gathering by visitors. The output of the simulation reveals indirect ecosystem effects by displaying the predic-tion of changes in other groups over time. Although a given model may not have ade-quate power to reveal all impacts, the simu-lation will reveal those that are most promi-nent. When a good model has been con-structed, the functional responses (direction and relative magnitude of changes) of groups can be revealing indicators for man-agers, while the indicated magnitudes of changes are taken as indicative only.  Ecospace is a new development of the Eco-path approach that enables resource manag-ers and scientists to simulate changes in spa-tial dynamics in response to trends of re-source use or management actions. It is a habitat-based approach in which compo-nents of the ecosystem achieve spatial dis-tributions according to movement rates and vulnerability to predators, which the user adjusts for each species in each habitat (de-fault rates are provided). For example, the effects of marine protected areas on fish stocks can be simulated in a particular situa-tion. Alternatively, the spatial and whole-ecosystem effects of habitat degradation can be investigated by comparing alternative models representing a change in habitat dis-tributions. The most appropriate type of management applications of the PWS model, as currently constructed, are those that have the potential to affect the Sound on its entire scale, such as fisheries management. The PWS model may not be appropriate for questions and projects that focus on a smaller scale, such as a single bay or fjord. Changes occurring on this smaller scale may be ecologically important at the local level, but it is unlikely that these changes would be detected using the broader PWS model. However, the PWS model could be used as a template for rapid construction of models of smaller areas, which could detect ecological impacts of local disturbances or protective measures. Future restoration and planning The general usefulness of the Ecopath ap-proach (with Ecosim and Ecospace) for fu-ture restoration and planning is evident by Figures 8-13. We suggest that these tools can not only aid in the scoping and devel-opment of assessment programs and in the analyses of their data, but they can also play an important role in future restoration and resource planning. Beyond the useful eco-system description using Ecopath, the spe-cific uses of the dynamic modelling routines discussed herein are to point to functional responses of components of the ecosystem based on the biological, trophic, and habitat information used in the model. It is these functional responses, rather than particular magnitude of responses, that are useful and reliable indicators of response to simulated disturbances, trends, or management actions. Most of all, the current and future usefulness and success of this approach depends on a functional collaborative process, such as the one engaged in during this project. 86     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Ecosystem Models as Carica-tures: the case of Prince William Sound Jennifer L. Ruesink University of Washington, Dept. of Zoology, Seattle It has become quite popular in ecological circles to decry the lost opportunity of effec-tively studying the Exxon Valdez oil spill. We worry that in the rush to point fingers and bill for damages, we failed to gather data that would allow any assessment of dy-namic responses to the spill and cleanup. Still, enormous efforts were expended in studies, many of which recorded abundance-- precisely the sort of data that Ecopath re-quires. This Ecopath project has collated an encyclopedic amount of information in one easily accessible form. At its best, it reveals not only point estimates of biomass, births, and deaths, but also their variability and un-certainty. The Ecopath model of Prince William Sound is a caricature, as are all models. But of course caricatures can be revealing, even though they are wrong. The values in this model are almost certainly incorrect: as an example, density estimates of intertidal epi-fauna varied by three orders of magnitude, even in a single habitat, and once these are averaged and corrected for Sound-wide habitat availability (for which there must be error in estimation, nor will it all have equivalent epifauna), the true tonnage can-not be known with any more certainty. The mechanisms in this model are likely to be incorrect as well, since the only mechanism presently used to represent interactions is direct consumption. It ignores the fact that consumers often do more than  skim produc-tion off of their prey; they can shift compo-sition to species with lower productivity and actually alter the production/biomass (P/B) of the group. Interactions also occur through direct competition or through habitat altera-tion; for instance, the organic structure of kelp and eelgrass beds is known to influence juvenile fish growth and mortality riska. Some of these mechanisms may occur within Ecopath 'boxes' and therefore their effects could be incorporated over short time frames. In a general sense, however, these errors in accuracy and mechanism only mat-ter if questions are asked that draw on as-pects of the model that are importantly wrong. It would be fooling, for instance, to use this caricature to set fishery quotas, or to predict all the effects of dredging a seagrass bed. The sort of question that can be asked of Ecopath models is, essentially, in which por-tion of the food web are dynamics most un-known? They may be unknown because no data exist (P/B and Q/B in particular were often copied from similar groups in other systems); or because confidence limits are large; or because it is unclear how details within groups should be expressed as an ag-gregate (although combined in a single group, many of the rockfish had essentially no diet or spatial overlap. Presumably con-sumption rates, conversion efficiencies, and productivities might also differ among spe-cies. What should the food web connections with this box be?); or because the ecotrophic efficiency is unlikely (way more is produced than used by the next trophic level, or vice versa). These unknown, uncertain, or unlikely portions of the food web bear addi-tional scrutiny. If more energy appears to be used than is available, for instance, further study of this nexus of interactions might re-veal new connections, resources, and unrec-ognized imports, not simply errors in esti-mation. In fact, it may prove more illuminating to leave these models unbalanced-- after all, there's no clear evidence that biomass and P/B are more poorly known than any other parameters, yet these tend to be adjusted because they cause fewer unwanted changes                                                         a A routine called "mediation" has recently been added to Ecosim which allows for non-feeding interactions between groups (see www.ecopath.org) (Editors). ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     87 elsewhere in the model. Balancing Ecopath models solves the puzzle one way out of a multidimensional space of possible ways, and we may be tempted to breathe a sigh of relief and say "Yes, that's how it must be." Complacent, we may be less inclined to search for biologically-based parameter re-visions and empirical evidence of what's actually going on in these interactions. It is possible to imagine that the system would indeed be 'out of balance' in the Eco-path sense for any number of reasons. 1. It is well-known that being spatially-explicit can change the outcome of interac-tions, even to the extent of whether or not a species persist. This issue may be remedied at a coarse scale in Ecospace. 2. Biomasses and productivities of PWS bi-ota change dramatically on a seasonal basis. In the current model, such fluctuations are simply averaged out-- 1 duck present for 3 winter months is just 0.25 ducks annually. On the other hand, the season of its resi-dence could coincide with a time when mus-sels are newly-settled and therefore not par-ticularly energetically valuable, but ex-tremely vulnerable to predation. Major changes also happen cyclically or catastro-phically on long time scales, but several models could be built to represent different conditions. 3. With some notable exceptions (salmon, bird predators), PWS is assumed to be a closed system, but many imbalances could be redressed through supplies or export of energy from outside the Sound. 4. There are no microbes in the model, de-spite recent recognition that the 'microbial loop' in pelagic systems can provide large proportions of energy to higher trophic lev-els. Even in benthic systems, much of the decomposition of macrophyte drift occurs by bacteria; and many benthic species con-sume bacterial films that grow on rock, soft sediment, or detritus. Including these food sources could substantially shift the basal production available to the system. On the other hand, including microbes might not have much effect on the relative strength of links currently incorporated in the model. Certainly, the issue could be explored. 5. Production is an amalgamation of proc-esses, including growth of surviving indi-viduals and recruitment of new individuals. It is fairly easy to accept that biomass growth will be a function of existing bio-mass (assuming constant size structure), but the relationship to recruitment is less clear-- fisheries biologists have been struggling with stock-recruitment curves for decades, and they are if nothing else exceedingly variable. One cohort of herring can domi-nate biomass for 10 years, because recruit-ment is so intermittent; there appears to be no way for Ecopath to link the fates of lar-vae in the zooplankton box to eventual re-cruitment events and adult populations. All that said, Ecopath is still a way to frame the question of whether these issues of scale, taxonomic focus, and parameterization actu-ally influence trophic flows. Incorporating these issues may not be necessary to under-stand trophic flows (though they should not be assumed unimportant), and certainly it complicates the model. The advantage of Ecopath over Ecosim is that mass-balance is a relatively simple mat-ter of solving simultaneous algebraic equa-tions. It is possible to determine exactly why the program gives the 'answer' it does. Eco-sim is less transparent, but it is dynamic and therefore able to generate predictions about the food web effects of perturbations. It would be interesting to know if these predic-tions are robust to assumptions about prey vulnerability. It would also be interesting to know if the predictions are testable; for in-stance, are any groups expected to move outside their range of natural variation in a time frame that would allow the change to be attributed to a particular event? Caricatures are wrong but useful as long as they're not asked to perform tasks involving portions that are importantly wrong. This Ecopath exercise has clearly been useful in pulling together disparate data about PWS. It has also usefully focussed attention on 88     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      uncertainties, variation, and trophic imbal-ances that may only be resolved by better studying and understanding the roles of spe-cies embedded within complicated webs of interaction. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     89 An annotated list of Alutiiq words relevant to modeling the Prince William Sound ecosystem   Dave Preikshot Fisheries Centre, UBC, Canada Jeff Leer Alaska Native Language Centre University of Alaska, Fairbanks The incorporation of traditional environ-mental knowledge (TEK) into environ-mental science has recently become a sub-ject of great interest. The cross fertilisation of TEK and western science could enrich both intellectual traditions. In the Alaskan context of rehabilitation efforts that have followed the Exxon Valdez oil spill in Prince William Sound (PWS) a large scien-tific research effort has begun in an area rich with TEK. To begin incorporating TEK into the scientific study of the PWS ecosystem (while also allowing natives to use scientific knowledge) a logical first step is to cata-logue local terms for the flora and fauna of PWS. This approach has been successfully applied to the Strait of Georgia ecosystem (Pauly et al. 1998) by using a catalogue of Saanich words for various fish species to help validate models of its potential historic state. In the PWS area a similar approach was deemed useful since TEK is often a valuable source of qualitative information as to the behaviour, location, and diet, of or-ganisms. Johannes (1981) provides a won-derful example of the intimacy of knowl-edge that can be possessed by a traditional sea faring people of the aquatic environment and its organisms. Practitioners of the ecological sciences have usually drawn the vast majority of their con-clusions on the investigation of quantitative  information. This has led to many notable achievements in the study of aquatic ecol-ogy, especially for modelling the population dynamics of single species. A relatively new approach however, has been to model whole ecosystems using approaches such as Eco-path (Christensen and Pauly 1993, and see other contributions in this volume). For gen-eral ecosystem modelling significant ad-vances may be achieved by including native knowledge holders in the scientific process. This is because many of the questions ad-dressed by ecosystem modelling are qualita-tive. TEK is also particularly helpful in the novel practice of modelling historic ecosys-tems. Often the TEK of the local community contains precise knowledge of species pre-sent (in both contemporary and historic con-texts), their diets and the seasonal fluctua-tions of populations. A particularly valuable aspect of such in-formation could be the determination of spe-cies that may have been present long ago, but have since been extirpated. By con-structing organism name inventories from different locales modelled ecosystems we may gain some insight as to distributions in time and space, population variations, and changes in diet. Such information may also be incorporated into relational databases such as Fishbase (Froese and Pauly 1998) to facilitate the determination of patterns. By such mechanisms local knowledge can be integrated into traditional scientific analysis. These databases should also help translate science into terms understandable by the local people themselves, since they allow the linking of biological concepts. An example of information on historical populations is the existence of a word for ‘mammoth’ in several languages of Arctic peoples from western Alaska through north-ern Canada and Greenland (Fortescue et al. 1994). The fact that so many languages have an established word, not a recent loan word, for this long since vanished animal suggests these languages have other valuable eco-logical information.  Equally interesting is the information that may be obtained from stories and oral his-tory. Although the Inuit from Greenland have no word for mammoth they do have one for a legendary six legged animal, the pronunciation of which is close to the word for mammoth in other Arctic languages (Fortescue et al. 1994). This is probably a construct from a cultural recollection of the 90     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      mammoth. It would not be surprising for the trunk of the beast to be later described as a leg. The inuit can be forgiven for thinking the creature had an even number of append-ages like most others, therefore describing it as six legged instead of possessing five ma-jor appendages. The language from which words for aquatic flora and fauna, and associated words were derived for this study was Alutiiq Alaskan Yupik (AAY), “…spoken in Alaska on the Shores of Prince William Sound, at the tip of the Kenai Peninsula, on Kodiak Island, and on the Alaska Peninsula.” (Fortescue et al. 1994). The language is, however, gener-ally referred to simply as Alutiiq (Leer 1978).   A few words must be said about the lan-guages of Arctic people. Generally these languages are referred to as ‘Eskimo’ lan-guages, although we recognise that some consider the use of the word Eskimo some-what inappropriate. This set of languages can be broken into three major groups; Inuit, spoken from the Seward Peninsula to Greenland; Yupik, spoken from the South shore of Norton Sound to PWS; Aleut, spo-ken in the Aleutian Islands. These languages are coastal, the interior of Alaska being dominated by Athabaskan languages. Other works have attempted to catalogue terms for aquatic species, such as McAllister et al. (1987) in their List of Inuktitut (Eskimo), French, English, and scientific names of ma-rine fishes of Arctic Canada. Such works, however, are not directly applicable to either the specifics of the language and ecosystem of PWS, and were therefore not used here. It will be necessary in the future to compare words in the languages of other studies like McAllister et al. (1987) to synthesise the knowledge of what are closely related lan-guages. Leer (1978) and Fortescue et al. (1994) split AAY into two dialects, Chugach (C) and Koniag (K). C was further split into two sub dialects: ‘Prince William Sound’, found in the eastern portion of the geographic distri-bution of the C dialect (CE) and ‘Kenai Pen-insula’ found in the west (CW). The K dia-lect was also split into two sub dialects; Ko-diak, found in the eastern portion of its geo-graphic distribution of K (KE), and Alaska Peninsula (KW), found in the west. When-ever possible the word from the CE subdia-lect was used for this study as it would be from people living closest to the ecosystem being modelled. If the CE  term was not available then words were used in the fol-lowing order of preference C>CW>K>KE>KW>AAY. See Figure 14, derived from Fortescue et al. (1994), for a guide to the locations of these dialects.    ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     91                          Due to limitations of the fonts available through the word processor used in this pro-ject many phonemes listed by Fortescue et al. (1994) have been approximated. This was not a difficulty with words from the list in Leer (1978) since he used phonemes with simple typewritten approximations. Follow-ing information from Pullum and Ladusaw (1986) the definitions of symbols were ob-tained so that a suitable symbol from the fonts available to the author could be used. The replacement symbols are as follows:  • An upside down lower case e, or schwa, meaning pronouncing the letter as a short e was denoted as ‘e’; • An l with a belt around it meaning slight aspiration on either side of the tongue when pronouncing l was denoted as ‘µ’; • An m with a small circle below the right hand arch, meaning slight aspiration through the nose when pronouncing the m was denoted as ‘mo’; • An n with a small circle underneath, meaning slight aspiration through the nose when pronouncing the n was de-noted as ‘no’; • An n with a small circle on the tip of the tail, meaning slight aspiration through the nose when pronouncing ‘ng’ was denoted as Îo; • A small capital r, as high as a lower case r, meaning heavy rolling of the letter when pronouncing r was denoted as ‘R’; • Verbs, which must be conjugated can be recognised by the Alutiiq transliteration having a dash on the end, thus ‘-’. Figure 14. Geographic range of Alutiiq Alaskan Yupik, its dialects and subdialects: CE is the Prince William Sound subdialect and CW is the Kenai Peninsula subdialect of Chugach. For the Koniag dialect KE represent the subdialect spoken in and around Kodiak while KW is the subdialect spoken on the Alaskan Peninsula. 92     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996       The following table was adapted from en-tries in Leer (1978), Fortescue et al. (1994), and the personal research notes of Jeff Leer. Entries that were taken from Leer (1978) and Fortescue et al. (1994) are cited as such. Entries from Jeff Leer’s research notes are denoted as ‘Leer notes’. In Fortescue et al. (1994) all phonemes were ordered alpha-betically, as in a dictionary. The structure of phonemes in Leer (1978) was different from Fortescue et al. (1994) and no typographical modifications were needed. Lastly, the words in Leer (1978) refer to C in general and are cited as such. Since this paper seeks to add terms from Alutiiq to a scientific data base it is sensible to order the terms into functional groups and then to alphabetise according to their English equivalents, thus making the list accessible to the largest pos-sible audience. The functional groups the terms were separated into are: general ani-mal terms, general bird terms, bird names, general fish terms, fish names, general mammal terms, mammal names, inverte-brates, and plants. The words are each de-scribed using the following format:     Format: English name / Alutiiq name / Language, Dialect, Subdialect / Remarks, if any ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     93 Animals, general  animal / uÎuÎsiq / KW / (Fortescue et al. 1994). animal / ungu’alaaq / CE / (Leer notes). animal, to take as game / pit’e- / C / (Leer notes). egg / pelisuq / C / (Leer notes). female / arnaqiitak / C / (Leer notes). game, caught / pitaq / C / (Leer notes). male / angusaluq / KW / (Leer notes). male / erilek / C / (Leer 1978). monster / cacalaa’ak / CE / (Leer notes). monster, lake / arwalaayak, ar’ulaayak / CW / (Leer notes). oil, from animals or plants / uquq / C / (Leer 1978). rib / cakia<R>aq / AAY / (Fortescue et al. 1994). The K subdialect also uses this word to refer to the rib of a boat. stomach / aqsaquq / C / (Fortescue et al. 1994). tusk , canine tooth / tuluRyaq, tuluRneq / AAY / (Fortescue et al. 1994).  Birds, general  backbone, upper part of / atankuyuk / CE / Leer notes.  beak, bill / cugg’eq / C / (Leer 1978). breastbone / qatek / CE / (Leer notes). crop / uniirwik / CE / (Leer notes). down feather / tenga’uk / CE / (Leer notes). egg, to lay / peksu- / C / (Leer notes). eggs, a complete set of, or batch, in a nest / naaneq / AAY / (Fortescue et al. 1994). eggs, to sit on / waa- / C / (Leer notes). feather / culuk / C / (Leer 1978). feather, tail / kingumik / CE / (Leer notes). feather, tail, long / culugpak / CE / (Leer notes). fledging bird / tengnerraq / C / (Leer notes). gizzard / aqsaqullnaa / CW / (Leer notes). nest / ungluq / C / (Leer notes). ptarmigan, crop of / pukuyaq / KW / (Leer notes). taking off and flapping wings against water sur-face / paa- / AAY / (Fortescue et al. 1994). to peck / pu’uγ / AAY / (Fortescue et al. 1994). wing / saqeq / C / (Leer notes). wings, to flap / saqiur- / C / (Leer notes). wishbone / agaq / CE / (Leer notes).  Bird names  albatross, black footed / iklliyayuusiq / CE / (Leer notes). i.e., Diomedea nigripes (Griggs 1997). albatross, short tailed / ungusarpak / CE / (Leer notes). i.e., Diomedea albatrus (Griggs 1997). auklet / akllegaq / CE / (Leer notes). There are four species of auklet in the PWS area; the rhinoceros auklet (Cerorhinca mono-cerata), Cassin’s auklet, (Ptychoramphus aleuticus) and the parakeet auktet (Cy-clorrhynchus psittacula) and the crested auklet (Aethia cristatella) (Griggs 1997). brant / kamouk / CE / (Fortescue et al. 1994). i.e., Branta bernicula.  chicken hawk / qecuwaliq / C / (Leer 1978). No synonym for ‘chicken hawk’ was found. There are four species of hawk which can be found in PWS; the red-tailed hawk (Buteo jamaicensis), rough-legged hawk (B. lagopus), goshawk (Accipiter gen-tilis), and sharp-shinned hawk (A. stria-tus) (Griggs 1997). coot / tekicehnquaq / CE / (Leer notes). i.e., Fu-lica americana. cormorant / agayuuq / C / (Leer notes). It seems unlikely that there is only one word for cormorant, since four species; the double crested cormorant (Phalacrocorax auri-tus), pelagic cormorant (P. pelagicus), Brandt’s cormorant (P. penicillatus), and the red faced cormorant (P. urile) are all found in PWS (Griggs 1997). cormorant, double breasted / agayuurpak / C / (Leer notes). Likely the double crested cormorant, Phalacrocorax auritus. cormorant, pelagic / uyalek / AAY / (Leer notes). i.e.,  Phalacrocorax pelagicus. cormorant, summer / plaatuugualek / CE / (Leer notes). Litrally means ‘one that has a ker-chief’ and therefore may refer to Brandt’s cormorant (Phalacrocorax penicillatus) which has a prominent blue chin or the red faced cormorant (P. urile) which has a red ‘mask’ on its face (Griggs, 1997). crane / tatellgaq / C / (Leer notes). i.e., Grus canadensis. crow / apalngaa’aq / CE / (Leer notes). i.e., the northwestern crow (Corvus caurinus), the only crow found in PWS (Griggs 1997). dipper / kui’im ayakutua / C / (Leer notes). i.e., Cinclus mexicanus. dowitcher, short-billed / kukukuaq / CW / (Leer notes), i.e., Limnodromus griseus. This word may also refer to the common snipe. duck / saquleq / C / (Leer 1978). This is a ge-neric word for any species of duck. 94     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      duck / ungusaq / CE / (Leer notes). This is a ge-neric word for any type of duck. duck, ‘eider-like’, small / extuk / AAY / (Fortes-cue et al. 1994). Whether this refers to ju-venile eiders of either species found in PWS, the common eider (Somateria mol-lissima) or the king eider (S. spectabilis), or other small duck like birds is unclear. This could refer to species such as scaups (Aythya), buffleheads and goldeneyes (Bucephala), or scoters (Melanitta) (Griggs 1997). There is also potential for steller’s eider (Polysticta stelleri) to be found in the PWS area. Note that extuk is the same word as that for a repeated sharp noise. duck, all white / nasqurtuli’aq / CE / (Leer notes). Perhaps a word for domesticated white ducks. duck, American widgeon / qacaaq / KW / (Leer notes). i.e., Anas americana. Note that this word may also be identified as the teal. duck, black scoter / sukumyaaq / CE / (Leer notes). i.e., Melanitta nigra. duck, bufflehead / nacallngaayak / C / (Leer notes). i.e., Bucephala albeola duck, canvasback / tengyuq / CE / (Leer notes). i.e., Aythya valisineria. This creature has also been identified as the ‘blue-billed’ duck by locals and has the synonym of egtuk. duck, eider / qayarriq / C / (Leer notes). There are two species of eider duck which fre-quent the waters of PWS; the king eider (Somateria spectabilis) and the common eider (S. mollissima) (Griggs 1997). This word is used in the KW sub-dialect to re-fer to the spectacled eider (S. fischeri). duck, eider, brown / qaanillqaacak / C / (Leer notes). May refer to the common eider (Somateria mollissima), but may also re-fer to the female eiders in general which are brown as in many species of duck. duck, falcated teal / kau’utaaq / CE / (Leer notes). i.e., Anas falcata. duck, gadwall / tengyunguaq / KW / (Leer notes). i.e., Anas strepera. Note that this word may also be identified as the green-winged teal (Anas crecca). duck, goldeneye / nasqurtuliq / CW / (Leer notes). There are two species of gold-eneyes found in PWS; the common gold-eneye (Bucephala clangula) and Barrow’s goldeneye (B. islandica). duck, goldeneye / qapugnaq / C / (Leer notes). There are two species of goldeneyes found in PWS; the common goldeneye (Bucephala clangula) and Barrow’s gold-eneye (B. islandica). duck, green-winged teal / apa’ariilnguq / CE / (Leer notes). i.e., Anas crecca. duck, green-winged teal / tengyunguaq / KW / (Leer notes). i.e., Anas crecca. Note that this word may also be identified as the gadwall (Anas strepera). duck, harlequin / qaingiaq / CE / (Leer notes). i.e., Histrionicus histrionicus. duck, king eider / qeÎaµek / AAY / (Fortescue et al. 1994). i.e., Somateria spectabilis. duck, mallard / ngillqitaq, nillqitaq / C / (Fortes-cue et al. 1994). i.e., Anas platyrhynchos. duck, mallard / seqtaq / CE / (Leer notes). i.e., Anas platyrhynchos. duck, oldsquaw / arrangkiluk / C / (Leer notes). i.e., Clangula hyemalis. duck, pintail / amutaarualek / C / (Leer notes). i.e., Anas acuta. duck, ring-necked / nasqurtuliq / KW / (Leer notes). i.e., Aythya collaris. duck, rock / ungunguasaaq / C / (Leer 1978). No synonym has been found for this word. duck, surf-scoter / tunuculek / CW / (Leer notes). i.e., Melanitta perspicillata.. duck, teal / qacaaq / KW / (Leer notes). There are two species of teal found in PWS; the green winged teal (Anas crecca), and blue-winged teal (Anas discors) (Griggs 1997). Note that this word may also be identified as the American widgeon (Anas americana). duck, white-winged scoter / gaalerualek / CE / (Leer notes). i.e., Melanitta fusca. eagle, bald / quckalaq / CE / (Leer notes). i.e., Haliaeetus leucocephalus. eagle, golden / anglluayuq / (Leer notes). i.e., Aquila chrysaetos. goose / tengmiaq / CE / (Leer notes). There is no distinction as to whether this word means the white fronted goose (Anser albifrons), the Canada goose (Branta canadensis), the brant (B. bernicula), or any two, or all of these species of goose which could be found in PWS (Griggs 1997). Note, too that the emperor goose (Chen canagica) could also be sighted in PWS. goose, Canada / laγiq / KW / (Fortescue et al. 1994). i.e., Branta canadensis. goose, white fronted / neqlleq / KE / (Leer notes). i.e Anser albifrons. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     95 goshawk / ulualek / CW / (Leer notes). i.e., Ac-cipiter gentilis. grebe, large / atatarpak / CE / (Leer notes). The red-necked grebe (Podiceps grisegena) is the largest of the two species of grebe found in PWS (Griggs 1997). grebe, red-necked / atatak / CW / (Leer notes). i.e., Podiceps grisegena. grebe, small / atataa’aq / CE / (Leer notes). The horned grebe (Podiceps auritus) is the smallest of the two species of grebe found in PWS (Griggs 1997). guillemot / cugaq / CE / (Leer notes). i.e., the pigeon guillemot (Cepphus grylle). gull, Bonaparte’s  / marayaaq / CW / (Leer notes). i.e., Larus philadelphia. This name is also used to refer to the long-tailed jaeger (Stercorarius longicaudus) and arctic tern (Sterna paradisaea). gull, glaucus / kukiswak / KW / (Fortescue et al. 1994). i.e., Larus hyperboreus. gull, herring / egyaaq / C / (Leer notes). i.e., La-rus argentatus. gull, sea / naru’aq / CE / (Leer notes). The mew gull (Larus canus), glaucus gull (L. hy-perboreus), herring gull (L. argentatus), glaucus winged gull (L. glaucescens) and Bonaparte’s gull (L. philadelphia) are all found in PWS (Griggs 1997). gyrfalcon / nerusicuulek / C / (Leer notes). i.e., Falco rusticolus. hawk, red-tailed / aarruliq / CE / (Leer notes). i.e., Buteo jamaicensis. hawk, rough-legged / qill’iq / CE / (Leer notes). i.e., Buteo lagopus. hawk, sharp-shinned / qecu’alia’aq / C / (Leer notes). i.e., Accipiter striatus. May also refer to the red-tailed hawk Buteo ja-maicensis. heron / yuaqurtuliq / CE / (Leer notes). i.e., great blue heron (Ardea herodias). jaeger, long-tailed / maraayaq / CW / (Leer notes). i.e., Stercorarius longicaudus. This word has also been identified as refering to Bonapartes’s gull (Larus philadelphia) and arctic tern (Sterna paradisaea). kingfisher, belted / nalu’alia’aq / CE / (Leer notes). i.e., Ceryle alcyon. kittiwake / qay’aqaaq / CE / (Leer notes). This probably refers to the black-legged kitti-wake ) Rissa tridactyla). kittiwake, red-legged / kiuksaa’aq / CW / (Leer notes). i.e., Rissa brevirostris. This word literally means ‘red-legged duck’ so there may be some confusion in the translation. loon / tuullek / C / (Leer notes). There are five species of loon in the PWS area: the red throated loon (Gavia stellata), the Pacific loon (G. pacifica), the common loon (G. immer), the yellow billed loon (G. adam-sii), and the arctic loon (G. arctica) (Griggs 1997). Informants on Kodiak is-land identified this as the word for the arctic loon (Leer notes). loon, red throated / qaqaaqaaq / K / (Fortescue et al. 1994). i.e., Gavia stellata. loon, small / quiriiq / CE / (Leer notes). Of the loon species found in PWS the two small-est are the red-throated loon (Gavia stel-lata) and the Pacific loon (G. pacifica) (Griggs 1997). magpie, summer / qallqanayuumiq / CE / (Leer notes). i.e., Pica pica. magpie, winter / manÕskia’aq / CE / (Leer notes). i.e., Pica pica. merganser, red-breasted / iisuuteklek / CE / (Leer notes). i.e., Mergus serrator. merganser, sawbill / paiq / C / (Leer notes). Probably refers to the common merganser (Mergus merganser), the merganser most seen in PWS (Griggs 1997). murre / allpaq / C / (Leer notes). Two species of murre are found in the PWS area; the common murre (Uria aalge) and the thick billed murre (U. lomvia) (Griggs 1997). murre, common / quanaaq / CE / (Leer notes). i.e., Uria aalge. The word literally means ‘thank-you’ and is a reference to the friendly demeanor of these birds. murrelet, marbled / taitui’aq / C / (Leer notes). i.e., Brachyramphus marmoratus. owl, great grey / eyiik / C / (Leer notes). i.e., Strix nebulosa. owl, great horned / yartuliiq / CE / (Leer notes). i.e., Bubo viginianus. owl, snowy / anipaq / CE / (Leer notes). i.e., Nyctea scandiaca. oystercatcher, black / kiggwikiaq / CE / (Leer notes). i.e., Haematopus bachmani. phalarope, northern / uqui’aq / CE / (Leer notes). There are two species of phalarope in PWS; the red necked phalarope (Phalaro-pus lobatus) and the red phalarope (P. fu-licaria) (Griggs 1997). ‘Northern’ phala-rope may be an Alaskan term for one of the two. puffin, horned / qilangaak / CE / (Leer notes). i.e., Fratercula corniculata. puffin, sea parrot / tunÎaq / KE / (Fortescue et al. 1994). It is unclear whether this refers to the horned puffin (Fratercula cornicu-96     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      lata) or the  the tufted puffin (F. cir-rhata). puffin, tufted / ngaq’ngaaq / C / (Leer 1978). i.e., Fratercula cirrhata. raven / apalngaaq / CE / (Leer notes). i.e., Cor-vus corax. sandpiper / ayakutaq / C / (Leer 1978). May also be called kui’im ayakutaa. There are sev-eral species of sandpiper native to PWS; the rock sandpiper (Calidris ptilocnemis), purple sandpiper (C. maritima), pectoral sandpiper (C. melanotos), Baird’s sandpi-per (C. baidii), western sandpiper (C. mauri), least sandpiper (C. minutilla), sanderling (C. alba), dunlin (C. alpina), solitary sandpiper (Tringa solitaris), and spotted sandpiper (Actitis macularia) (Greer 1997). sawbill duck / paiq / C / (Leer 1978). This may refer to either of two mergansers; red-breasted (Mergus serrator) or common (M. merganser) found in PWS. These ducks have obvious serations on their bills. snipe / kulickiiq / CE / (Leer 1978). i.e., the common snipe (Gallinago gallinago). De-rived from Russian. swan / uquirpak / CE / (Leer notes). It is not specified whether this refers to the trum-peter swan (Cygnus buccinator) or the tundra swan (C. columbianus) both of which occur in PWS. It may also refer to the whooper swan (C. cygnus) although it is now an extremely rare species (Griggs 1997). swan / saqulegpak / KW / (Leer notes). May refer to the whooper swan (Cygnus cyg-nus). tern, arctic / ayusaq / C / (Leer 1978). i.e., Sterna paradisaea. tern, common / teki’aq / KW / (Leer notes). i.e., Sterna hirundo, which does not normally extend as far northwest as PWS (Griggs 1997).  Fish, general  cod egg / mac’utak / CE / (Leer notes). fish / iqalluk / C / (Leer 1978).  fish cloaca / qurwikusaaq / KW / (Leer notes). fish pew for tossing fish / ipuun / AAY / (Fortes-cue et al. 1994). fish eggs / lluu’ak / CE / (Leer notes). fish eggs soaked in fresh water / qaRmit / AAY  / (Fortescue et al. 1994). This word also means ‘to be crunchy’, a textural phe-nomenon occuring when fish eggs are put in fresh water. fish eggs, aged, added to cooked salmon eggs (fish-egg cheese) / piinaq / C / (Fortescue et al. 1994). fish eggs, membrane containing / puγneq / AAY / (Fortescue et al. 1994). fish eggs, mixed with oil, mashed potatoes and other ingredients / akutaq / C / (Leer 1978). fish fin / suluksuk / KW / (Leer notes). fish fin, anal / pamyursuun / C / (Leer notes). fish fin, caudal, paddle, oar  / angua’un / AAY  / (Leer notes). Note the overlap of fin and oar. fish fin, dorsal / culugsuun / C / (Leer notes). fish fin, pelvic / saqiu’um / CW / (Leer notes). fish gill / pacik / C / (Leer notes). fish head, aged / uqsuq / CE / (Fortescue et al. 1994). fish head, bones in bioled / mat’ruat / CE / (Leer notes).  fish head, cod / iicumaaq / CE / (Leer notes). fish meat, drying / kinertaq / AAY / (Leer notes). fish milt / napasaaq / CE / (Leer notes). fish pie / piËuk, piluk / C / (Leer 1978). From the Russian word pirok. fish rack / initaarwik / C / (Leer 1978). fish scale / qugleq / CE / (Leer notes). fish skin, / ami’aq / C / (Leer notes). May also refer to dried fish skin. fish slime / nuayaaq / C / (Leer notes). fish tip / tuqlluq / C / (Leer notes). fish trap / taluyaq / AAY / (Fortescue et al. 1994). fish, a cut piece of / kep’aq / KW / (Leer notes). fish, aged / cin’aq / AAY / (Leer notes). fish, aged, aged fish eggs / qulunguaq / CE / (Leer notes). fish, boiled / egaapiaq / C / (Leer 1978). fish, boiled, half dry / uumataq / C / (Leer 1978). fish, cut / sege- / C / (Leer notes). fish, dark meat under skin, fish kidney / qet’aq / CE / (Leer notes). fish, dead, found along a river after spawning / urullciq / CE / (Leer notes). fish, dried / mingciq / CE / (Leer notes). fish, dried / tamuuq / AAY / (Fortescue et al. 1994). fish, head cartilage / tatangquq / AAY / (Leer notes). fish, old / aakanaq / C / (Leer 1978). fish, raw / qasaq / AAY / (Fortescue et al. 1994). fish, raw, to eat / qasaR- / AAY / (Fortescue et al. 1994). fish, salt / sulunaq / C / (Leer 1978). ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     97 fish, salted, smoked / palik / CW / (Leer 1978). cf the AAY term ‘to smoke fish’, puyuqe-. fish, skeleton cut for drying / ataneq / C / (Leer 1978). fish, swim bladder / pagaaciq / CE / (Leer notes). fish, to smoke / puyuqe- / AAY / (Fortescue et al. 1994). fish, to spawn / qarya- / CW / (Leer notes). minnow / napi’aq / CE / (Leer notes). salmon, rear half / taggwi / ? / (Leer notes). salmon, male pink with hump on back / amaqataq / KW / (Leer notes). salmon, moldy and dying after spawning / uu-kanaqicuk / CE / (Leer notes). salmon, turned red after entering fresh water / nariqaaq / CE (Leer notes). salmon, with hump after entering fresh water / qutnguq / C / (Leer notes). tom cod egg / arhmaasuuk / C / (Leer notes).  Fish names  bass, black / tukuq / C / (Leer 1978). Clemens and Wilby (1961) state that ‘black bass’ is a term used to refer to two different spe-cies of rockfish (genus Sebastes) on the West Coast of North America, the blue rockfish (Sebastes mystinus) and the black rockfish (Sebastes melanops). bass, sea / tilpuuk / CW / (Leer notes). i.e., white sea bass (Atractoscion nobilis). capelin / cikeq / AAY / (Fortescue et al. 1994). i.e., Mallotus villosus. cod / amutaq / C / (Leer 1978) i.e., Pacific cod (Gadus macrocephalus). cod, arctic / atgiaq / CW / (Leer notes). i.e., Bo-reogadus saida. cod, kelp / culugpua’ak / CE / (Leer notes). No scientific name could be associated with this fish. cod, white / quuguuk / CW / (Leer notes). No scientific name could be associated with this fish. cod-like fish, long with big eyes / iituliq / ? / (Leer notes). No scientific name could be associated with this fish. eel / quguutnaq / C / (Leer notes). It is not speci-fied what type of eel this refers to. It may also refer to any eel like fish.  eulachon / cikeq / KW / (Leer notes). i.e., Thaleichthys pacificus. flounder, rough-skinned / ggagtuliq / C / (Leer 1978). i.e., the starry flounder (Platich-thys stellatus), which has rough scales. flounder, smooth-skinned / matuqulluk / CE / (Leer notes). Many species of flounders and flounder like fish (family Pleuronec-tidae) have smooth skin. It is uncertain whether this word refers to one in particu-lar, or all non rough-skinned flounders in general. flounder, starry / ur’auk / CE / (Leer notes). i.e., Platichthys stellatus. hake, Pacific / rririliq / C / (Leer notes). i.e., Merluccius productus. This word may also refer to the whiting (Theragra chal-cogrammus), which is closely related to Pacific hake halibut / sagiq / C / (Leer notes). i.e., Hippoglos-sus stenolepis. herring / iqalluarpak / C / (Leer 1978). i.e., Clu-pea pallasi. Irish lord / nyangtaaq / CE / (Leer notes). i.e., the red Irish lord (Hemilepidotus hemilepidotus). The brown Irish lord (H. spinosus) does not normally range as far north as Alaska (Clemens and Wilby 1961). An English synonym for this fish is bullhead. It should, however, be noted that the word el’ista in the CE subdialect is the name for a larger variety of bull-head (Leer notes). Therefore, nyangtaaq may refer to the brown Irish lord and el’ista may refer to the red Irish lord since the latter is usually much larger than the former. lumpfish / amrruq / C / (Leer notes). i.e., the spiny lumpsucker (Eumicrotremus orbis). pike / qalru / KW / (Leer notes). i.e., Esox lucius. salmon / iqaµuk / AAY / (Fortescue et al. 1994). Note that this word may also be used to refer to other fish. salmon or trout fry / ilaRnaq / KE / (Fortescue et al. 1994). salmon, chinook (king) / iqallugpak / CE / (Leer notes). i.e., Oncorhynchus tshawytscha. salmon, chum (dog) / alngartuliq / CE / (Leer notes) i.e., Oncorhynchus keta. salmon, chum (dog), old, after spawning / kaÎitnoeq / K / (Fortescue et al. 1994). i.e., Oncorhynchus keta. salmon, cutthroat trout / talaa’ik / CE / (Fortes-cue et al. 1994). i.e., Oncorhynchus clarki. salmon, pink / amarturpiaq / CE / (Leer notes). i.e., Oncorhynchus. gorbuscha. salmon, red (sockeye) / niklliq / C / (Leer 1978). i.e., Oncorhynchus nerka. salmon, silver (coho) / caayuaq / CE / (Fortescue et al. 1994). i.e., Oncorhynchus kisutch. 98     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      salmon, steelhead / mayu’artaq / C / (Leer notes). i.e., Oncorhynchus mykiss. sculpin, red-bellied / asirnaq / CW / (Leer notes). Refers to a species locally known as the red-bellied sculpin. No scientific name could be found for this fish. There are many sculpin species (family Cotti-dae) in PWS. sculpin, yellow / kala’aq / CW / (Leer notes). Refers to a species locally known as the ‘yellow sculpin’. No scientific name could be found for this fish. There are many sculpin species (family Cottidae) in PWS. shark / qaacaq / C / (Leer 1978). The three most commonly observed sharks in PWS are; the spiny dogfish (Squalus acanthias), salmon shark (Lamna ditropis), and six-gill shark (Hexanchus griseus) (see Hul-bert, this vol.) skate / sagirniilnguq / C / (Leer notes). This could refer to any of the skate species (family rajidae) found in PWS smelt, boreal / iqalluaq / KW / (Leer notes). Pre-sumably refers to capelin (Mallotus villo-sus), although other smelt (Osmeridae) do reside in PWS. snapper, red / ushmaq / CE / (Leer notes). i.e., Sebastodes ruberrimus. sole / tasaayaq / KW / (Leer notes) / This could refer to any of the numerous species of sole and flounder (family Pleuronectidae) inhabiting PWS. stickleback / cukilrua’ak / CE / (Leer notes). i.e., Gasterosteus aculeatus. tomcod / taaqatak / CE / (Leer notes). i.e., Mi-crogadus proximus. trout / saagua’aq / CE / (Leer notes). Any mem-ber of the family Salmonidae. trout, hook-nosed / curlluk / KE / (Leer notes). No scientific names has been as-sociated as of yet with this fish. trout, spotted / giigaq / KW / (Leer notes). Both the Dolly Varden (Salvelinus malma) and brook trout (S. fontinalis) have prominent spots. whiting / rririliq / C / (Leer notes). i.e., Theragra chalcogrammus. This word might also re-fer to the Pacific hake (Merluccius pro-ductus), a close relation of whiting.  Mammals, general  ambergris / kulamiim miryaa, kulamiim qu-laq’aa / CE / (Leer notes). Literally, ‘whale vomit’. baleen / negarkaq / C / (Leer notes). fin, dorsal / puguun / CE / (Leer notes). flipper, front / it’ga’aq / KW / (Leer notes). flipper, tail / it’alaq / C / (Leer notes). food in stomach or intestines / imanaq / AAY / (Fortescue et al. 1994). This word is re-lated to similar words in other Eskimo languages for a species of mollusk eaten by walrus. fur / amiq / C / (Leer 1978). gut, blown up / suplluaq / CE / (Leer notes). Usually from a bear or sea lion. meat, between ribs and fat of a seal / qiak / KW / (Leer notes). membrane, covering seal gut / katu’arneq / CE / (Leer notes). oil, rendered / egneq / AAY / (Leer notes). Mostly derived from seals. oil, to fry in, render oil by frying blubber / cua-taaR / C / (Fortescue et al. 1994). Mostly derived from seals. oil, to render / ege- / C / (Leer notes). pelt, skin side / cata / CW / (Fortescue et al. 1994). porpoise skin / mangtak / C / (Leer notes). sea lion flipper, gristly layer underneath / mau-nak / C / (Leer notes). seal bladder / meq’artaq / CE / (Leer notes). Used as a buoy, or to carry fresh water. seal fat / usulkiiq / AAY / (Fortescue et al. 1994). seal fetus / imlauq / KW / (Fortescue et al. 1994). seal head / aalisuuk / KW / (Leer notes). seal hide, dehaired by hanging over hot rocks in a sauna / ulikuq / CW / (Leer notes). seal hide, dehaired by stretching and scraping, used for a kayak cover / nengugtaq / C / (Leer notes). seal hide, old, used for various purposes / cimyaq / C / Leer notes). seal meat between ribs and fat / qiak / KW / (Leer notes). seal oil / blubber / uquq / AAY / (Fortescue et al. 1994). seal, pelvis bone / mak’atestaaq / C / (Leer notes). Used to play a kind of divining game sealskin float / awataq / KW / (Fortescue et al. 1994). sealskin, to take blubber off of / qapagte- / AAY / (Leer notes). sinews / qikarlluk / KW / (Leer notes). Usually derived from caribou and sea lion and used for sewing. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     99 skin, tanned, from young aninmal / qatgwiaq / CW / (Leer notes). Used as underwear. walrus tusk, ivory / tugka’aq / KW / (Leer notes). i.e., the tusk of Odobenus rosma-rus. whale tail / caqrwik / K / (Leer notes).  Mammal names  bear, black / tan’erliq / C / (Leer notes). i.e., Ur-sus americanus. bear, brown / laqlaq / C / (Leer notes). i.e., the grizzly bear (Ursus arctos). bear, grizzly / taquka’aq / K / (Leer notes). i.e., Ursus arctos. dolphin / qaaniq / C / (Leer 1978) i.e., Pacific white sided dolphin (Lagenorhynchus obliquidens). fur seal / aataak / C / (Leer 1978). i.e., Cal-lorhinus ursinus. otter, sea / ikam’aq / C / (Leer 1978). i.e., Enhy-dra lutris. porpoise / cilpiq / C / (Leer notes). This word refers to a porpoise with a long (dorsal?) fin. When diving four points from its body break the surface. porpoise / mangaq / C / (Leer notes). It is not specified whether this refers to the harbor porpoise (Phocoena phocoena) or Dall’s porpoise (Phocoenoides dalli) both of which are found in PWS (Hill et al. 1997). sea lion / wiinaq / C / (Leer notes). There is no distinction whether this word refers to the steller sea lion (Eumetopias jubatus) or the northern fur seal (Callorhinus ursinus) both of which may be found in PWS (Hill et al. 1996). seal / qaigyaq / C / (Leer notes). It seems most likely that this word refers to the harbor seal (Phoca vitulina), but it can also refer to other seals. The northern elephant seal (Mirounga angustirostris) is the only other likely to be seen in PWS (Hill et al. 1997). seal, small that does not grow / nainguaq / KW / (Leer notes). No scientific name was founf for this animal. seal, spotted / alngalck / C / (Leer notes). i.e., Phoca largha. Not now resident in PWS (Hill et al. 1997). walrus / asguq / CE / (Leer notes). i.e., Odobenus rosmarus. whale, bowhead / ar’uq  / CW / (Leer notes). i.e., Balaena mysticetus, a species not resident in PWS (Hill et al. 1997). whale, humpback / qenÎulek / CE / (Fortescue et al. 1994). i.e., Megaptera novaeagliae. Fortescue et al. (1994) note that this might also refer to whales in general, as they only cite one author for the hump-back whale definition. whale, killer / arlluk / C / (Leer notes). i.e., Or-cinus orca. whale, large type / taksugpak / CE / (Leer notes). The most common large whales that would be seen in or around PWS are the gray whale (Eschrichtius robustus), the humpback whale (Megaptera novaean-gliae), the fin whale (Balaenoptera phy-salis), and the northern right whale (Eubalaena glacialis) (Hill et al. 1997). The fin whale is the largest and the term may refer specifically to it, but this is not specified. The whale is said to be narow and long, one informant said sixty feet (twenty meters) long, with teeth, suggest-ing it may be the sperm whale (Physeter catodon). whale, minke / mangarniiq / C / (Leer notes). i.e., Balaenoptera acutorostrata. whale, sperm / kulamak / CE / (Leer notes). i.e., Physeter catodon. This word can also re-fer to any whale. whale, white / asi’arnaq, anaqarnaq / C / (Leer notes). i.e., the beluga whale (Delphinap-terus leucas). whale, with ‘carved’ breast / uniinalek / CE / (Leer notes). No scientific name was found for this whale.  Invertebrates  anemone, sea / sanaqusak / C / (Leer 1978). i.e., members of the class Anthozoa.  barnacle / qauq / C / (Leer notes). i.e., a species of either genus Semibalanus or Balanus. chiton, gumboot / urriitaq / C / (Leer 1978). i.e., Cryptochiton stelleri. chiton, ladyslipper / uriitarpak / C / (Leer 1978). i.e., an unidentified member of the class Polyplacophora. clam / salaq / C / (Leer 1978) i.e., members of the class Bivalvia. clam, geoduck / salarpak / CE / (Leer notes). i.e., Panope abrupta. clam, razor / cingtaataq / C / (Leer notes). i.e., Siliqua patula. clam, red neck / tuuqaatiq / CE / (Leer notes). No scientific name was found for this 100     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      species. clam, long-necked / alirualek / CE / (Leer notes). No scientific name was found for this species. clam, small high water / set’alek / CE / (Leer notes). No scientific name was found for this species. cockle / taugtaaq / C / (Leer notes). i.e., Clino-cardium nuttalli. crab / yual’aak / CE / (Leer notes) i.e., members of the order Decapoda, infraorder Brachyura. crab, dungeness / canipgaq / CE / (Leer notes). i.e. Cancer magister. crab, tanner / pupsuleryu’alq, pupsuleryua’ak / CE / (Leer notes). i.e., Chionoecetes bairdi. cucumber, sea / kingugpak / CE / (Leer notes). i.e., members of the class Holothuroidea. tusk shell / iwiluryaaq / C / (Leer notes). i.e., Dentalium pretiosum. flea, sand / petgeryaaq / AAY / (Fortescue et al. 1994). i.e., Orchestia traskiana. invertebrate, marine / imaam kingua / CW / (Leer notes). Literally means ‘sea bug’. jellyfish / qaacek / C / (Leer 1978). i.e., members of class Scyphozoa, order Semaeo-stomeae. limpet, Chinaman’s hat / melungqucak / CE / (Leer notes). i.e., belongs to the family Acmaeidae. mussel / amyak / C / (Leer notes). Since there is no further clarification, the word might refer to large individuals of either or both species, Mytilus edulis and M. trossulus, found in Alaska (Foster 1997). mussel, big brown / melugyaq / KW / (Leer notes). Apparently refers to the genus Modiolus. octopus / amikuk / C / (Leer notes). i.e., members of the family Octopodidae. oyster / qailim matutii / CE / (Leer notes). i.e. members of the genus Crassostrea. sand flea / qumitgaq / c / (Leer 1978). i.e., mem-bers of the order Amphipoda sea star / agsiq / C / ‘Starfish’ in Fortescue et al. (1994). i.e., members of the class Aster-oidea. sea star / agyaruaq / CW / (Leer 1978). Note the difference with the word in Fortescue et al. (1994); both are used in CW. This one is a newly coined word meaning ‘some-thing like a star’. sea urchin / uutuk / AAY / (Fortescue et al. 1994). i.e., members of the class Echin-oidea. sea worm, black / anaqiitak / C / (Leer notes). No scientific name was found for this species. shrimp / petgeryaarpak / CE / (Leer notes). i.e. a member of the class Decapoda. snail / ipuk / CW / (Leer notes). i.e., members of either order Mesogastropoda or order Neogastropoda. snail, large / ipuullquq / C / (Leer notes).  snail, coffee / kauglaq / C / (Leer notes). No scientific name was found for this snail. squid / amikurniilnguq / CE / (Leer notes). i.e., members of the family Loliginidae.  Plants / Protists  algae / aqayak / KE / (Fortescue et al. 1994). driftwood / pukilaaq / C / (Leer 1978). driftwood, bark used for fire / ketaq / C / (Leer 1978). driftwood, small piece / camRu(q) / AAY / (Fortescue et al. 1994). eelgrass / cuula’ik / CE / (Leer notes). kelp, brown / set’alek / CE / (Leer notes). A brown kelp with pencil-like marks and two leaves on each side of the stem. The stem is eaten. kelp, brown, heavy, flat, wide / cimyaruaq / CE / (Leer notes). Perhaps a member of the family Laminariaceae. kelp, bulb / aqlluurteshnaq / CE / (Leer notes). No synonym was found for this species. kelp, bull / qalinguq / C / (Leer notes) i.e., Nere-ocystis leutkeana. kelp, bull, head / nasquluk / CW / (Leer notes). kelp, bull, tail (whip) of / nuakataq / C / (Leer notes). Often used for fishing line. kelp, green, large round head / meq’artaq / CE / (Leer notes). No synonym was found for this species. kelp, long stringy / arlluguaq / CE / (Leer notes). Has gas bladders strung out like buoys at intervals. kelp, sheet / kapuustaaruaq / C / (Leer notes). A large, green coloured species, the name of which was derived from the Russian word for cabbage. Perhaps a member of the family Laminariaceae. kelp, streamer of / sel’aq / CW / (Leer notes). log, drift / tep’aq / AAY / (Leer notes). plant, eaten by swans from lake bottoms / qer-qaq, rrertaq / AAY / (Leer notes). seaweed, branched in fingers / ata’ik / CE / (Leer ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     101notes). Possibly refers to members of the family Codiaceae. seaweed, bulbous / iituq / C / (Leer notes). A kelp with hole in the blade, perhaps a ref-erence to the appearannce of the blade of reproducing Nereocystis luetkeana indi-viduals. seaweed, edible (dulse) / caqallqaq / C / (Leer 1978). seaweed, edible (dulse), large / caqallqarpak / CE / (Leer notes). seaweed, fucus / ellquaq / CW / (Leer notes). i.e., Fucus gardneri. seaweed, long,  fishline / nemeRyaq / AAY / (Fortescue et al. 1994). seaweed, hair / nuaruaq / CE / (Leer notes). No synonym was found for this species. seaweed, red / nepuaq / CE / (Leer notes). weeds, water / ney’aq / CE / (Leer notes).  DISCUSSION One of the most striking features of this word list is the contrast between the groups of organisms for which names appear. For instance the coverage of names for mam-mals and birds is comparatively rich whereas the detail of names for inverte-brates, and plants is less informative. Such varying detail can be explained by two proc-esses, the first a result of the linguists who recorded Alutiiq terms, the second a result of the Alutiiq people who acted as infor-mants. In obtaining terms for a species linguists must know something of the classification of those animals they are seeking to define. It is not surprising, therefore, that there is more details for bird and mammal species. These are the most familiar aquatic species to lay people and non-specialists and the easiest to observe from land. It should also be noted that of the fish named, the most detailed information relates to economically important species, which, again would be more familiar to someone lacking a special-ists’ knowledge of local organisms. Problems may also arise in translating some terms. For example, Fortescue et al. (1994) base their dictionary on many sources which originally sought to translate ‘Eskimo’ terms to European languages other than English, including Danish, Russian, German, and French. This implies a larger scope for con-fusion in obtaining common names for ani-mal species than in English alone, an already rich language for common names. For ex-ample, Oncorhynchus tshawytscha can be referred to as ‘king salmon’, ‘tyee salmon’, ‘spring salmon’, ‘quinnat salmon’, or ‘chi-nook salmon’ depending solely on the locale on the west coast of North America. There-fore, in the final unification of all the sources to the dictionary there is much room for confusion between common European names. Problems of ascribing too much detail may occur such as in the popular myth of ‘Es-kimo’ languages having many terms for snow. This myth has been effectively re-futed by Martin (1986). However, as Pullum (1991) points out the story is still held to be true by many, including linguists and scien-tists. The story has its basis in the Worfian hypothesis that a people’s environment helped shape the richness of their language. Therefore, for northern peoples it was logi-cal that there must be many words for snow. Numbers of words for snow in Eskimo lan-guages have been variously reported as high as 400 (see Pullum 1991). The myth was also fed by paternalistic urban views of the type of society Inuit peoples must have and has been magnified through time. As for the informants, they too act as a filter of information. As Berlin (1992) points out that informants quickly become aware of the level of expertise of the person studying their language. Given this understanding, informants will provide more detail where they believe the greater understanding lay. So if the informants felt the linguist was un-familiar with invertebrates of the area, they may quite logically omit detail that might be wasted on an observer incapable of distin-guishing the different species. A further layer of filtering may occur in cases where people having knowledge of certain types of animals were simply not questioned. For example the division of la-bour in traditional societies may be accord-102     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      ing to class, family or sex. Often the upper class may have exclusive rights to hunt large game animals. Another common phenome-non is women being the people most often responsible for collecting invertebrates and plants from the nearshore environment. It is easy to imagine that linguists studying a given community may in all likelihood spend the majority of their time with male hunters. By this mechanism there would be an increased likelihood of informant-driven filtering occurring. Members of the less talked to groups would likely see the linguist as having very little knowledge or interest in the species they may be asked to give infor-mation on. Berlin (1992) gives an example of such a mechanism for women of the Aguaruna who were unwilling to provide knowledge they had of manioc species (a plant) since male questioners were felt to be ‘ignorant’ on the topic. We can see from these issues a need to bridge the gap between TEK and traditional science. Most important to bridgeing this gap is a clear understanding of what differ-ent people mean when they use a particular word. This paper represents such a step for PWS by cross referencing animals and con-cepts with their scientific equivalents. We have attempted to minimise difficulties, con-fusion and assumptions by the long term contact Jeff Leer has had with communities of Alutiiq speakers throughout his career as a linguist. His continuing contact with speakers in the communities from which the words in this study were used allows for a greater degree of precision and accuracy than could be achieved by simply consulting a dictionary. Without such interpretation local TEK could be misrepresented, as either not specific or overly specific. For example, a people who fish will have much to say about the habits of the species they target. However, they may have somewhat less to say about the habits of other organisms which they do not use. Such distinctions are useful in deter-mining what knowledge TEK and science can most effectively exchange. Cultures do classify organisms in roughly similar fash-ion, but these classifications reflect the rela-tive abundance of the organisms they en-counter (Berlin 1992). Thus, by accurately identifying the words in two languages and the organisms they refer to, we can begin to establish the bridges required for cross vali-dation of TEK and science.  ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     103 ACKNOWLEDGEMENTS We thank the Exxon Valdez Oil Spill Trus-tee Council for support of these efforts to synthesize PWS ecosystem information (project no. BAA 98330). We are grateful for the guidance and encouragement of res-toration program scientists A. Gunther and B. Spies, as well as B. Wright of the Na-tional Oceanic and Atmospheric Administra-tion and S. Senner and P. Mundy of the Council's restoration program office. C. Walters and V. Christensen provided and refinined the modelling tools and software needed for this work.  We thank J. Ruesink for her thorough review and contribution of an inciteful commentary on the utiliity of this approach. Most of all, we thank the con-tributors to this volume and other project leaders who saw the potential of this ap-proach to synthesize PWS ecosystem knowledge. We thank S. Pimm and B. Pow-ell for their complementary analysis of the earlier versions of the PWS model. Their suggestions, scrutiny, and attention to detail was a positive contribution to the final model. C. Young, with the contributions of A. Poon, did a supurb job managing the pro-duction of the CD ROM, which includes this report, the PWS model, other Alaska models and documentation, the Ecopath with Eco-sim software and much more. Finally, the editors thank A. Beattie, L. Vidal, N. New-lands, J. Dalsgaard, and D. Preikshot for contributing advice, knowledge, and en-thusiasm, and L. Thompson for valuable suggestions and support.  104     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      LITERATURE CITED Aasen, O. 1963. Length and growth of the por-beagle (Lamna nasus) in the north-west Atlantic. Fisk. Dir. Skr. (Ser. Havunders.) 13(6):20-37 Agler, B. A., and S. J. Kendall.  1997. Marine bird and sea otter population abundance of Prince William Sound, Alaska: trends following the T/V Exxon Valdez Oil Spill, 1989-96. Exxon Valdez oil spill restora-tion final reports, Restoration Project 96159. U.S. Fish and Wildl. Serv., An-chorage, Alaska.  Alaska Department of Natural Resources and US Fish and Wildlife Service. 1991. The Exxon Valdez oil spill natural resource damage assessment and restoration. A re-port on oiling to environmentally sensi-tive shoreline. Report to the Exxon Valdez Oil Spill Trustee Council. Ambrose, R. E., and K. E. Riddle. 1988. Popula-tion dispersal turnover, and migration of Alaska peregrines, p. 677-684. In: T. J. Cade, J. H. Enderson, C. G. Thelander, and C. M. White (eds.). Peregrine Falcon populations, their management and recov-ery. The Peregrine Fund, Boise, ID.  Arrhenius, F. and S. Hansson. 1993. Food Con-sumption of larval, young and adult her-ring and sprat in the Baltic Sea. Marine Ecology Progress Series 96:125-137. Ashmole, N. P. 1971. Sea bird ecology and the marine environment, p. 224-286. In: D. S. Farner, H. R. King, and K. C. Parkes (eds.), Avian biology, vol. 1. Academic Press, New York. Ashwell-Erickson, S. M. 1981. The energy cost of free existence for Bering Sea harbor and spotted seals. Ph. D. Thesis. Univ. Alaska, Fairbanks. 209 p. Ashwell-Erickson, S. M., and R. Elsner 1981. The energy cost of free existence for Ber-ing Sea harbor and spotted seals, p. 869-899. In: D. W. Hood and J. A. Calder, (eds.) The eastern Bering Sea shelf: oceanography and resources. Vol. 2, U.S. Dep. Commer., NOAA, Off. Mar. Pollut. Assess., Juneau, Alaska.  Bailey, K.M., R.D. Brodeur, and A.B. Hollowed. 1996. Cohort survival patterns of walleye pollock, Theragra chalcogramma, in Shelikof Strait, Alaska: a critical factor analysis. Fish. Oceanogr. 5(1): 179-188. Baird, Robin. W. 1994. Foraging behaviour and ecology of transient killer whales (Or-cinus orca). Ph.D., Simon Fraser Univer-sity. 157 p.  Barnes, R.D. 1987. Invertebrate Zoology, 5th Edition. Saunders College Publishing, Philadelphia, 893 p. Barraclough, W.E. and J.D. Fulton. 1967.  Num-ber, size composition, and food of larval and juvenile fish caught with a two-boat surface trawl in the Strait of Georgia, July, 1966. Fish. Res. Board Can. Man. Rep. Ser. 940: 82 p. Barraclough, W.E. and J.D. Fulton. 1968.  Num-ber, size composition, and food of larval and juvenile fish caught with a two-boat surface trawl in the Strait of Georgia, July, 1966. Fish. Res. Board Can. Man. Rep. Ser. 940: 82 p. Barrett-Lennard, L.G. 1992. Echolocation in wild killer whales (Orcinus orca). M.S., University of British Columbia. 74 p.  Barrett-Lennard, L.G., K. Heise, E. Saulitis, G. Ellis, and C. Matkin. 1995. The impact of killer whale predation on Steller sea lion populations in British Columbia and Alaska. North Pacific Universities Marine Mammal Research Consortium, Univer-sity of British Columbia, Vancouver, B.C., unpubl. rep., 66 p. Beissinger, S. R. 1995. Population trends of the Marbled Murrelet projected from demo-graphic analyses, p. 385-393. In: C. J. Ralph, G. L. Hunt, M. G. Raphael, and J. F. Piatt (eds.), Ecology and conservation of the Marbled Murrelet. Pacific South-west Res. Sta., Albany, CA. Berlin, B. 1992. Ethnobiological classification: principles of categorization of plants and animals in traditional societies. Princeton, Princeton Univ. Press. 335 p. Bigg. M. A. 1969. The harbour seal in British Columbia. Fish. Res. Board Can. Bull. 172. 33 p. Birt-Friesen, V. L., W. A. Montevecchi, D. K. Cairns, and S. A. Macko. 1989. Activity-specific metabolic rates of free-living Northern Gannets and other seabirds. Ecology. 70:357-367. Blagoderov, A.I. 1994. Seasonal distribution and some notes on the biology of salmon shark (Lamna ditropis) in the Northwest-ern Pacific Ocean. J. Ichthy. 34(2):115-121 ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     105Bodkin,  James L. and Mark S. Udevitz. 1996. 1993 Trial aerial survey of sea otters in Prince William Sound, Alaska.  Exxon Valdez Oil Spill Restoration Project Final Report (Restoration Project 93043-2) . National Biological Service, Anchorage, Alaska. Bodkin, James L. and Mark S. Udevitz. 1991. Development of sea otter survey tech-niques. In: Ballachey, B.E., J.L. Bodkin and D. Burns (eds.). Assessment of the magnitude, extent, and duration of oil spill impacts on sea otter populations in Alaska; Marine Mammals Study No. 6. Natural resources damage assessment status report, November 19, 1991. U.S. Fish and Wildlife Service, Anchorage, Alaska. Bowman, T. D., P. F. Schempf, and J. A. Berna-towicz. 1993. Effects of the Exxon Valdez oil spill on Bald Eagles. Exxon Valdez oil spill State/Federal Natural Resource Damage Assessment Final Report. U.S. Fish and Wildl. Serv., Anchorage, Alaska.  Bowyer, R.T., J.W. Testa, J.B. Faro, C.C. Schwartz and J.B. Browning. 1994. Changes in diets of river otters in Prince William Sound, Alaska: effects of the Exxon Valdez  oil spill. Can. J. Zool. 72:970-976. Braddock, J. F., J. E. Lindstrom, T. R. Yeager, and B. T. Brown. 1996. Patterns of mi-crobial activity in oiled and unoiled sedi-ments in Prince William Sound. Pages 94-108 In: Rice, S. D., R. B.Spies, D. A. Wolfe, and B. A. Wright (eds.). Proceed-ings of the Exxon Valdez Oil Spill Sym-posium. American Fisheries Society Symposium 18, Bethesda. Bradford, M.J. 1995. Comparative review of Pacific salmon survival rates. Can. J. Fish. Aquat. Sci. 52:1327-1338. Bright, D.B. 1959. The occurrence and food of the sleeper shark Somniosus pacificus in a central Alaska bay. Copeia 1:76-77. Brown, J., A. Colling, D. Park, J. Phillips, D. Rothery, J. Wright. 1989. The Ocean Ba-sins: Their Structure and Evolution. Per-magon Press, Inc., Elmsford, New York. 171 p. Calkins, D.G. 1978. Feeding behavior and major prey species of the sea otter, Enhydra lu-tris, in Montague Strait, Prince William Sound, Alaska. Fishery Bulletin  76:125-131. Cass, A. J., R. J. Beamish, and G. A. McFarlane. 1990. Lingcod (Ophiodon elongatus). Can. Spec. Pub. Fish. Aquat. Sci. No. 109. Department of Fisheries and Oceans, Ottawa. Castro, J.I. 1983. The sharks of North American waters. College Station: Texas A&M University Press. Christensen, V. 1996. Balancing the Alaska gyre model, p. 32-36. In: D. Pauly and V. Christensen (eds.)  Mass-Balance Models of North-eastern Pacific Ecosystems. Fisheries Centre Research Report 4 (1). Christensen, V. and D. Pauly., Editors. 1993. Trophic models of aquatic ecosystems. International Centre for Living Aquatic Resources Management, ICLARM Conf. Proc. 26. Manila, Philippines, 390 p. Clemens, W. and G. Wilby. 1961. Fishes of the Pacific coast of Canada. Bulletin No. 68 (second edition), Fish. Res. Board Can. 443 p. CNF. 1998. Analysis of the Management Situa-tion: Summary. Chugach National Forest, 3301 C Street Anchorage, Alaska 99503-3998, (907) 271-2560. Compagno, L.J. 1984. FAO Species catalogue, Vol 4, parts 1 & 2 Sharks of the World. United Nations Development Program.  Cooney, R. T. 1997. Sound Ecosystem Assess-ment (SEA): A Science Plan for the Res-toration of Injured Species in Prince Wil-liam Sound, Alaska. Exxon Valdez Oil Spill Restoration Project Annual Report (Restoration Project 96320), Institute of Marine Science, University of Alaska Fairbanks, Alaska. 99775-7220.  Cooney, R.T. 1993. A theoretical evaluation of the carrying capacity of Prince William Sound, Alaska, for juvenile Pacific salmon. Fisheries Research 18:77-87. Costa, D. P. 1982. Energy, nitrogen and electro-lyte flux and sea water drinking in the sea otter, Enhydra lutris. Physiological Zool-ogy. 55:34. Crovetto, A., J.  Lamilla, and G.  Pequeno. 1992.  Lissodelphis peronii, Lacepede 1804 (Delphinidae, Cetacea) within the stom-ach contents of a sleeping shark, Som-niosus cf. pacificus, Bigelow and Shroe-der 1994, in Chilean waters.  Marine Mammal Science 8(3):312-314. Dahlheim, M, A. York, R. Towell, J. Waite, and J. Breiwick. Harbor porpoise (Phocoena phocoena) abundance in Alaska: Bristol 106     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Bay to Southeast Alaska, 1991-1993. Submitted to Marine Mammal Science.  Dalsgaard, J. and D. Pauly. 1997. Preliminary mass-balance model of Prince William Sound, Alaska, for the pre-spill period, 1980-1989. Fisheries Centre Research Report 5(2), 34 p. Dalsgaard, J., S.S. Wallace, S. Salas, D. Preik-shot. 1998. Back to the future: recon-structing the Strait of Georgia ecosystem. In: Pauly, D. T. Pitcher and D. Preikshot (eds.). Back to the future: reconstucting the Strait of Georgia ecosystem. Fisheries Centre Research Reports 6(5). Damker, D. M. 1977. Initial zooplankton inves-tigations in Prince William Sound, Gulf of Alaska, and Lower Cook Inlet. Envi-ronmental Assessment of the Alaska Con-tinental Shelf, Annual Reports of the Principal Investigators for the Year End-ing 1977. 10(Receptors-fish, littoral, ben-thos):137-274. Dean, T. A., M. S. Stekoll, and R. O. Smith. 1996a. Kelps and oil: The effects of the Exxon Valdez oil spill on subtidal algae. Pages 412-423 In: Rice, S. D., R. B.Spies, D. A. Wolfe, and B. A. Wright (eds.). Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18, Bethesda. Dean, T. A., M. S. Stekoll, S. C. Jewett, R. O. Smith and J. E. Hose. Eelgrass (Zostera marina L.) in Prince William Sound, Alaska: Effects of the Exxon Valdez oil spill. Marine Pollution Bulletin (In press 1998). Dean, T. A., S. C. Jewett, D. R. Laur, and R. O. Smith. 1996b. Injury to epibenthic inver-tebrates resulting from the Exxon Valdez oil spill. Pages 424-439 In:: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (eds.). Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18, Be-thesda. del Hoyo, J., A. Elliott, and J. Sargatal (eds.). 1996. Handbook of the birds of the world, Vol. 3, hoatzin to auks. Lynx Edicions, Barcelona.  DeSanto, T. L., and S. K. Nelson. 1995. Com-parative reproductive ecology of the auks (family Alcidae) with emphasis on the Marbled Murrelet, p. 33-47. In: C. J. Ralph, G. L. Hunt, M. G. Raphael, and J. F. Piatt (eds.), Ecology and conservation of the Marbled Murrelet. Pacific South-west Res. Sta., Albany, CA. Dolphin W.F. Prey densities and foraging of humpback whales, Megaptera novaean-gliae. Experientia 43  Doroff, A.M. and J.L. Bodkin. 1994. Sea otter foraging behavior and hydrocarbon levels in prey. p. 193-208 In: T. Loughlin (ed.). Marine mammals and the Exxon Valdez. Academic Press. San Diego, CA.  Duffy, D. C. 1997. APEX Project: Alaska Preda-tor Ecosystem Experiment in Prince Wil-liam Sound and the Gulf of Alaska. Exxon Valdez Oil Spill Restoration Project An-nual Report (Restoration Project 96163 A-Q), Alaska Natural Heritage Program and Department of Biology, University of Alaska Anchorage, Alaska. Dunning Jr., J. B. 1993. CRC handbook of avian body masses. CRC press. Boca Raton. Emerson, S., C. Stump, P. M. Grootes, M. Stuiver, G. W. Farwell, F. H. Schmidt. 1987. Estimates of degradable organic carbon in deep-sea surface sediments from 14C concentrations. Nature 329 (6134):51-53. Estes, J. A., and D.O. Duggins. 1995. Sea otters and kelp forests in Alaska: Generality and variation in a community ecological para-digm. Ecological Monographs 65(1):75-100. Estes, J. A., M. T. Tinker, T. M. Williams, and D. F. Doak. 1998. Killer whale predation on sea otters linking oceanic and near-shore ecosystems. Science 282:473-476.  Estes, J. A., N. S. Smith, and J. F. Palmisano. 1974. Sea otter predation and community organization in the western Aleutian Is-lands. Ecology 59:822-833. EVOSTC. 1994. Exxon Valdez Oil Spill Restora-tion Plan. Exxon Valdez Oil Spill Trustee Council. Anchorage, Alaska. EVOSTC. 1996. Exxon Valdez Oil Spill Restora-tion Plan: Update on Injured Resources and Services. Exxon Valdez Oil Spill Trustee Council. Anchorage, Alaska. Fahl K., and R. Stein. 1997. Modern organic carbon deposition in the Laptev sea and the adjacent continental slope: Surface water productivity vs. terrigenous input. Organic Geochemistry 26(5-6):379-390. Feder, H. M. and S. C. Jewett. 1986. The sub-tidal benthos. p. 347-396 In: D. W. Hood and S. T. Zimmerman (eds.). The Gulf of Alaska: physical environment and bio-ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     107 logical resources. U.S. Department of Commerce, NOAA, Washington, D.C. OCS study, MMS 86-0095. Feder, H. M. and S. C. Jewett. 1988. The subtidal benthos. pp. 165-202 In D. G. Shaw and M. J. Hameedi (Eds.) Envi-ronmental Studies in Port Valdez, Alaska: a basis for management. Lec-ture Notes on Coastal and Estuarine Studies 24, Springer-Verlag, New York.  Feder, H. M. and T. M. Blanchard. 1998. The deep benthos of Prince William Sound, Alaska, 16 months after the Exxon Valdez oil spill. Marine Pollution Bulletin 36(2):118-130. Feder, H.M. 1995. Injury to deep benthos. Final Report to Exxon Valdez Oil spill Trustee Council.  70 p. + Appendices. Feder, H.M. and S.C. Jewett. 1988. The subtidal benthos, p. 165-202. In:on Coastal and Estuarine Studies, Vol. 24. Springer-Verlag, Berlin. Feder, H.M., A.S. Naidu, J.M. Hameedi, S.C. Jewett, and W.R. Johnson. 1989. The Chukchi Sea continental shelf: benthos-environmental interactions. OCSEAP Fi-nal Report (RU 687). Ford, J. K. B. 1994. Killer Whales: the natural history and genealogy of Orchinus orca in British Columbia and Washington State. University of Washington Press, Seattle. 102 p. Ford, J. K. B. 1999. Transients: mammal-hunting whales of British Columbia, Washington, and Southeastern Alaska. UBC Press, Vancouver. 96 p. Fortescue, M., S. Jacobsen, and L. Kaplan. 1994. Comparative Eskimo dictionary with Aleut cognates. Fairbanks, Alaska Native Language Center, University of Alaska. 614 p. Foster, N. 1997. The molluscan fisheries of Alaska. P. 131-144, in C.L. MacKenzie, Jr., V.G. Burrell, Jr., A. Rosenfield, and W.L. Hobart (eds.). 1997. The history, present condition, and future of the mol-luscan fisheries of North and Central America and Europe, Vol. 2, Pacific Coast and supplemental topics. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 128, 217 p. Foster, N. 1997. The molluscan fisheries of Alaska. p. 131-144, in C.L. MacKenzie, Jr., V.G. Burrell, Jr., A. Rosenfield, and W.L. Hobart (eds.). 1997. The history, present condition, and future of the mol-luscan fisheries of North and Central America and Europe, Vol. 2, Pacific Coast and supplemental topics. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 128. Froese, R. and D. Pauly, eds. 1998. FishBase 98: concepts, design, and data sources. ICLARM, Manila, Philippines. 293 p. Gaedke, U. 1995. A comparison of whole-community and ecosystem approaches (biomass size distributions, food web analysis, network analysis, simulation models) to study the structure, function and regulation of pelagic food webs. J. Plankton Res. 17(6):1273-1305. Garshelis, D. L. and C. B. Johnson. 1999. Otter-eating orcas. Science 283:175d Gayanilo, F. C., Jr. and D. Pauly, (eds.) FAO-ICLARM stock assessment tooks. (FiSAT). Reference manual. FAO Com-puterized Information Series (Fisheries ). No. 8. Rome, FAO. 262 p. Goering, J. J., W. E. Shields, and C. J. Patton. 1973. Primary production. p. 253-277 In D. W. Hood, W. E. Shields and E. J. Kel-ley, (eds.). Environmental Studies of Port Valdez. Inst. Mar. Sci., Univ. Alaska Fairbanks. Occas. Publ. 3. Goudie, R. I., and C. D. Ankney. 1986. Body size, activity budgets, and diets of sea ducks wintering in Newfoundland. Ecol-ogy 67:1475-1482. Goudie, R. I., and P. C. Ryan. 1991. Diets and morphology of digestive organs of five species of sea ducks wintering in New-foundland. J. Yamashina Inst. Ornithol. 22:1-8. Grebmeier, J. M. and J. P. Barry. 1991. The in-fluence of oceanographic processes on pe-lagic-benthic coupling in polar regions: A benthic perspective. J. Mar. Syst. 2(3-4):495-518.  Griggs, J. 1997. All the birds of North America. New York, HarperCollins. 174 p. Grubb, T. G., and R. J. Hensel. 1978. Food hab-its of nesting Bald Eagles on Kodiak Is-land, Alaska. Murrelet. 59:70-72. Guénette, S. 1996. Macrobenthos. p. 65-67 In: D. Pauly and V. Christensen (eds.). Mass-balance models of North-eastern Pacific Ecosystems. Fisheries Centre Research Report 4(1).  Hart, J.L. 1973. Pacific Fishes of Canada. Bull. Fish. Res. Board Canada 180:44-47. 108     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Hatch, S. A., and M. A. Hatch. 1990. Compo-nents of breeding productivity in a marine bird community: key factors and concor-dance. Can. J. Zool. 68:1680-1690. Hatfield, B. B., D. Marks, M. T. Tinker, K. Nolan and J. Peirce. 1998. Attacks on sea otters by killer whales. Marine Mammal Science 14(4):888-894. Hedges, J.I., R.G. Keil, and R. Benner. 1997. What happens to terrestrial organic matter in the ocean? Organic Geochemistry 27(5-6):195-212. Highsmith, R. C., M. S. Stekoll, P. G. van Tamelen, S. M. Saupe, T. L. Rucker, and L. Deysher. 1995. Herring Bay monitor-ing and restoration studies. Report to the Exxon Oil Spill Trustee Council, Anchor-age, Alaska. Highsmith, R. C., M. S. Stekoll, W. E. Barber, L. McDonald, D. Strickland, and W. P. Erickson. 1994. Comprehensive assess-ment of coastal habitat Study No. 1A. Re-port to Exxon Valdez Oil Spill Trustee Council, Anchorage, Alaska. Hilborn, R. 1996. Detecting population impacts from oil spills: A comparison of method-ologies. Pages 639-644 In:: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (eds.). Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18, Be-thesda. Hill, P.S., D.P. DeMaster, and R.J. Small. 1997. Alaska marine mammal stock assess-ments, 1996. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-AFSC-78, 150 p. Hobbs, R. C., and J. A. Lerczak, 1993. Abun-dance of Pacific white-sided dolphin and Dall’s porpoise in Alaska estimated from sightings in the North Pacific Ocean and the Bering Sea during 1987 through 1991. Annual report to the MMPA Assessment Program, Office of Protected Resources, NMFS, NOAA, Silver Spring, Maryland Hoberg, M.K. 1986. A numerical analysis of the benthic infauna of three bays in Prince William Sound, Alaska. M. S. thesis,  Humboldt State University, Arcada, Cali-fornia. 153 p. Holland-Bartels, L. 1997. Mechanisms of Impact and Potential Recovery of Nearshore Ver-tebrate Predators. Exxon Valdez Oil Spill Restoration Project Annual Report (Res-toration Project 96025). Alaska Science Center, Biological Resources Division, Anchorage, Alaska. Hollowed, A.B., B.A. Megrey, and E. Brown. 1993. Walleye pollock, p. 1, 1-1, 54. In: Groundfish planning team (eds.), Stock Assessment and Fishery Evaluation Re-port for the Groundfish Resources of the Gulf of Alaska as Project for 1994. North Pacific Fishery Management Council, Anchorage, Alaska.  Hoskin, C.M. 1977. Macrobenthos from three fjords in western Prince William Sound, Alaska. Report R77-1, Institute of Marine Science, Fairbanks, Alaska, 28 p. Houghton, J. P., Lees, D. C., and Driskell, W. B. 1993. Evaluation of the condition of Prince William Sound shorelines follow-ing the Exxon Valdez oil spill and subse-quent shoreline treatment. Biological Monitoring Survey, Vol. 2, NOAA Tech-nical Memorandum NOS ORCA 73. NOAA, Seattle, Washington. Howe, A. L., G. Fidler, A. E. Bingham, and M. J. Mills. 1996. Harvest, catch, and participa-tion in Alaska sport fisheries during 1995. Alaska Department of Fish and Game, Fishery Data Series No. 96-32, Anchor-age.  Howe, A. L., G. Fidler, and M. J. Mills. 1995. Harvest, catch, and participation in Alaska sport fisheries during 1994. Alaska Department of Fish and Game, Fishery Data Series No. 95-24, Anchor-age.  Howe, A. L., G. Fidler, C. Olnes, A. E. Bing-ham, and M. J. Mills. 1997. Harvest, catch, and participation in Alaska sport fisheries during 1996. Alaska Department of Fish and Game, Fishery Data Series No. 97-29, Anchorage. Huato, L. 1996. Salmon in the Ocean, p. 21-23. In: D. Pauly and V. Christensen (eds.)  Mass-Balance Models of North-eastern Pacific Ecosystems. Fisheries Centre Research Re-port 4 (1). Isleib, P., and B. Kessel. 1973. Birds of the North Gulf Coast - Prince William Sound Region, Alaska. Biol. Pap. Univ. Alaska 14. Iverson, S. J., K. J. Frost, L. F. Lawry. 1997. Fatty acid signatures reveal fine scale structure of foraging distribution of har-bor seals and their prey in Prince William Sound, Alaska. Marine Ecology Progress Series 151: 255-271. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     109 Jacob, K.H. 1986. Seismicity, Tectonics, and Geohazards of the Gulf of Alaska Re-gions. p. 145-184 in D.W. Hood and S.T. Zimmerman, (eds.). The Gulf of Alaska: Physical Environment and Biological Re-sources. U.S. Dept. of Commerce, Na-tional Oceanic and Atmospheric Admini-stration.  Jefferson, T. A. 1988. Phocoenoides dalli. p. 1-7 In: Mammalian Species No 319. The American Society of Marine Mammalo-gists. Jewett, S. C. &and T. A. Dean. 1997. The effects of the Exxon Valdez oil spill on shallow subtidal eelgrass communities in Prince William Sound, Alaska 1990-95. Final Report to the Exxon Valdez Oil Spill Trustee Council. 110 p. + Appendices. Jewett, S. C. and T. A. Dean. 1995. The effects of the Exxon Valdez oil spill on eelgrass communities in Prince William Sound, Alaska 1990-95. Report to the Exxon Val-dez Oil Spill Trustee Council. Anchorage, Alaska. Jewett, S.C., T.A. Dean,  R.O. Smith, L.J. Hal-dorson, D. Laur, M. Stekoll and L. McDonald. 1995. The effects of the Exxon Valdez oil spill on shallow subtidal communities in Prince William Sound, Alaska 1989-93. Final Report to the Exxon Valdez Oil Spill Trustee Council, Anchorage, Alaska. 178 p. + Appendices. Johannes, R. E. 1981. Words of the lagoon: fish-ing and marine lore in the Palau district of Micronesia Berkely, Univ. of California Press. 245 p. Johnsgard, P. A. 1993. Cormorants, darters and pelicans of the world. Smithsonian Insti-tution Press. Washington. Karpov, K.A. and G.M. Cailliet. 1978. Feeding dynamics of Loligo opalescens. p. 45-66 In: Reckseik, C.W. and Frey, H.W. Bio-logical, Oceanographic, and Acoustic as-pects of the market squid, Loligo opales-cens Berry. California Department of Fish and Game, Fish Bulletin 169.  Kastelein, R. A., J. Hardeman and H. Boer. 1997. Food consumption and body weight of harbor porpoises (Phocoena pho-coena). p. 217-233. In: Read, A. J., P. R. Weipkema and P. E. Nachtigall (eds.) The biology of the harbor porpoise. De Spil Publishers, The Netherlands. Keeble, J. 1991. Out of the Channel: The Exxon Valdez Oil Spill in Prince William Sound. HarperCollins, New York, 290 p. Kirsch, J. 1997. Acoustic biomass estimate of adult walleye pollock in Prince William Sound, Alaska, in winter 1997. Final re-port to the Alaska Dept. of Fish and Game, Prince William Sound Science Center, Cordova, Alaska. Kline, T. C., Jr. and D. Pauly. 1998. Cross-validation of trophic level estimates from a mass-balance model of Prince William Sound using 15N/14N data. pp. 693-702 In: F. Funk, T. J. Quinn II, J. Heifetz, J. N. Ianelli, J. E. Powers, J. F. Schweigert, P. J. Sullivan, C. I. Zhang (Eds.) Fishery Stock Assessment Models, University of Alaska Sea Grant, Fairbanks. Klosiewski, S. P., and K. K. Laing. 1994. Marine bird populations of Prince William Sound, Alaska, before and after the Exxon Valdez oil spill. NRDA Bird Study Num-ber 2. Unpubl. Rep., U.S. Fish and Wildl. Serv., Anchorage, Alaska. Koehl, P. S., T. C. Rothe, and D. V. Derksen. 1982. Winter food habits of Barrow's goldeneyes in southeast Alaska. p. 1-5 In: Nettleship, D. N., G. A. Sanger, and P. F. Springer (eds.). Marine birds: their feed-ing ecology and commercial fisheries re-lationships. Proc. Pacific Seabird Group Symp. Can. Wildl. Serv. Special Pub. Kvitek, R. G. and J. S. Oliver. 1992. Influence of sea otters on soft-bottom prey communi-ties in Southeast Alaska. Marine Ecology Progress Series 82(2):103-113.  Kvitek, R. G. and J. S. Oliver. A.R. de Gange, B.S. Anderson. 1992. Changes in Alaskan soft-bottom prey communities along a gradient in sea otter predation. Ecology 73(2):413-428. Lang, G.M., P.A. Livingston, R. Pacunski, J. Parkhurst, M.S Yang.  1991.  Groundfish food habits and predation on commer-cially important prey species in the east-ern Bering Sea from 1984-1986. NOAA Tech. Memo. NMFS F/NWC-207, 240 p.  Laur, D. and L. Haldorson. 1996. Coastal habitat studies: the effect of the Exxon Valdez oil spill on shallow subtidal fishes in Prince William Sound, p. 659-670. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (eds.). Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18, Be-thesda. Leatherwood, S., R. R. Reeves, W. F. Perrin, and W. E. Evans. 1982. Whales, dolphins, and porpoises of the eastern North Pacific and 110     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      adjacent arctic waters. NOAA Technical Report, NMFS Circular 444.  Leer, J. 1978. Nanwalegmiut Paluwigmiut-Ilu Nupugnerit: Conversational Alutiiq Dic-tionary, Kenai Peninsula Alutiiq. Alaska Native Language Centre, University of Alaska, Fairbanks. 306 p. Lees, D. C. and W. B. Driskell. 1981. Appendix IV. Investigations on shallow subtidal habitats and assemblages in lower Cook Inlet. In. Distribution, abundance, com-munity structure and trophic relationships of the nearshore benthos of Cook Inlet. H.M. Feder, A.J. Paul, M.K. Hoberg, and S.C. Jewett, authors. Environmental As-sessment of the Alaska Continental Shelf, Final Reports of Principal Investigators 14(Biological Studies):419-610. Lenihan, H.S., J.S. Oliver, and M.A. Stephenson. 1990. Changes in hard bottom communi-ties related to boat mooring and tributyltin in San Diego Bay: A natural experiment. Marine Ecology Progress Series 60(1-2):147-159.  Livingston, P.A. 1996. Pacific Cod and Sable-fish, p. 41-42. In: D. Pauly and V. Chris-tensen (eds.). Mass-balance models of North-eastern Pacific Ecosystems. Fisher-ies Centre Research Report 4(1).  Loentief, W.W. 1951. The structure of the U.S. Economy, 2nd ed. Oxford University Press, New York. Loughlin, T.R. (ed.). 1994. Marine Mammals and the Exxon Valdez. Academic Press, Inc. San Diego. 395 p. Love, M. 1996. Probably more than you want to know about the fishes of the Pacific coast, 2nd ed. Really Big Press, Santa Barbara, CA. 381 p. Luning, K. 1990. Pages 355-370. Seaweeds: Their Environment, Biogeography, and Ecophysiology. John Wiley and Sons, Inc., New York. Macdonald, R.W., S.M. Solomon, R. E. Cran-ston, H.E. Welch, M.B. Yunker, and C. Gobail 1998. A sediment and organic car-bon budget for the Canadian Beaufort shelf. Marine Geology 144(4):255-273.  Martell, S. 1999. Incorporating seasonality into Ecopath. FishBytes 5(6):1&3. Published by the Fisheries Centre, University of British Columbia, Vancouver. Martin, L. 1986. Eskimo words for snow: a case study in the genesis and decay of an an-thropological example. American An-thropologist 88 (2): 418-423. McAllister, D.E., V. Legendre, and J.G. Hunter. 1987. List of Inuktitut (Eskimo), French and scientific names of marine fishes of Arctic Canada. Can. Man. Rept. Fish. Aquat. Sci. 1932. 106p. McConnaughey, T. 1978. Ecosystems naturally labeled with carbon-13: applications to the study of consumer food-webs. Thesis presented to University of Alaska, Fair-banks, Alaska. McConnaughey, T. and C.P. McRoy. 1979. C 13 label identifies eelgrass (Zostera marina) carbon in an Alaskan estuarine food web. Marine Biology 53:263-269. McEvoy, A.F. 1986. The Fishermen’s Problem: Ecology and Law in the California Fisher-ies. Cambridge University Press, New York, NY, USA, 368 p. McFarlane, G.A. and R.J. Beamish. 1983. Pre-liminary observations on the juvenile bi-ology of sablefish (Anoplopoma fimbria) in waters off the west-coast of Canada, p. 119-136. In: Proceedings of the Interna-tional Sablefish Symposium. Alaska Sea Grant Rep. 83-8. McRoy, C.P. 1988. Natural and anthropogenic disturbance at the ecosystem level, Chap-ter 13. In: D.G. Shaw, M.J. Hameedi (eds.). Environmental Studies in Port Valdez, Alaska. Lecture notes on Coastal and Estuarine Studies 24, Springer-Verlag, Berlin. McRoy, C.P. and T. Wyllie Echeverria. 1990. Food Webs from selected species from Prince William Sound. Technical Report, Institute of Marine Sciences, University of Alaska, Fairbanks, 46 p. Menge, B. A. 1995. Indirect effects in marine rocky intertidal interaction webs: Patterns and importance. Ecol. Monogr. 65(1):21-74. Meyer, S.C. 1995. Recreational halibut fishery statistics for southcentral Alaska (Area 3A), 1994. A report to the International Pacific Halibut Commission. Alaska De-partment of Fish and Game, Special Pub-lication No. 96-1, Anchorage. Meyer, S.C. Composition and Biomass of the Recreational Rockfish Sebastes Harvest in South-central Alaska, 1992-1995. Alaska Department of Fish and Game, Data Series Report, Anchorage, (In prep). ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     111 Morstad, S., D. Sharp, J. Wilcock and J. John-son. 1997. Prince William Sound Man-agement Area 1996. Annual Finfish Man-agement Report. Alaska Department of Fish and Game. Commercial Fisheries Management and Development Division, Central Region, Anchorage, Alaska.  Nagasawa, K. 1998. Predation by salmon sharks (Lamna ditropis) on Pacific salmon (On-chorhynchus spp.) in the North Pacific Ocean. Pac. Anadr. Fish Comm. Bull No. 1: 419-433 Naidu, A. S. and L. H. Klein. 1988. Sedi-mentation processes. pp. 69-91 In D. G. Shaw and M. J. Hameedi (Eds.) Environmental Studies in Port Valdez, Alaska: a basis for management. Lec-ture Notes on Coastal and Estuarine Studies 24, Springer-Verlag, New York.  NAS. 1970. The great Alaska earthquake of 1964: Human Ecology. National Acad-emy of Sciences, Washington, DC, USA.  Neave, F., and M.G. Hanavan. 1960. Seasonal distribution of some epipelagic fishes in the Gulf of  Alaska region. J. Fish Res. Board. Canada, 17(2):221-233 Nelson, R. W. 1977. Behavioral ecology of coastal peregrines (Falco peregrinus pealei). Ph.D. Thesis, University of Cal-gary, Calgary, Alberta. NMFS. 1993. Prince William Sound Trawl as-sessment, Final Report State\Federal Natural Resource Damage Assessment Fish/Shellfish Project no. 18, National Marine Fisheries Service, Alaska Fisher-ies Science Center. NMFS. 1996. Our living oceans: report on the status of U.S. living marine resources, 1995. U.S. Dep. Commer. NOAA Tech. Memo. NMFS-F/SPO-19, 160 p. NPFMC. 1995. Stock assessment and fishery evaluation report for the groundfish re-sources of the gulf of Alaska. Compiled by the Plan Team for the Groundfish Fisheries of the Gulf of Alaska for the North Pacific Fisheries Management Council, Anchorage, Alaska. NPFMC. 1997. Stock assessment and fishery evaluation report for the groundfish re-sources of the gulf of Alaska. Compiled by the Plan Team for the Groundfish Fisheries of the Gulf of Alaska for the North Pacific Fisheries Management Council, Anchorage, Alaska. NRC. 1996. The Bering Sea Ecosystem. National Research Council, National Acadamy Press, Washington, DC, USA, 307 p. O’Clair, C. E. and S. D. Rice. 1985. Depression of feeding and growth rate of the seastar Evasterias troschelii during long-term exposure to the water-soluble fraction of crude oil. Marine Biology 84:331-340. Okey, T. A. 1997. Sediment flushing observa-tions, earthquake slumping, and benthic community changes in Monterey Canyon head. Continental Shelf Research 17(8):877-897. Okey, T. A. and D. Pauly. 1999. A mass-balanced model of trophic flows in Prince William Sound: De-compartmentalizing ecosystem knowledge. pp. 621-635 In: S. Keller (ed.). Ecosystem approaches for fisheries management. University of Alaska Sea Grant, Fairbanks. O'Reilly, J. E. 1985. Carbon flow in a coastal marine bottom community. Dissertation Abstracts International Part B: The Sci-ences & Engineering 45(11):214 pp. Orsi, J.A., J.M. Murphy, and D.G. Mortensen. 1998. Survey of juvenile salmon in the marine waters of southeast Alaska, May-August 1998. (NPAFC Doc. 346) Auke Bay Laboratory, Alaska Fisheries Science Center, National Marine Fisheries Ser-vice, NOAA, Juneau, Alaska, 26 p. Ostrand, W. D., and J. M. Maniscalco. 1996. Seabird/forage fish interactions, p. B1-B25. In: D. C. Duffy (ed.). APEX: Alaska predator ecosystem experiment, Exxon Valdez oil spill restoration project annual report, Exxon Valdez Oil Spill Trustee Council, Anchorage, Alaska. Ostrand, W. D., K. O. Coyle, G. S. Drew, J. M. Maniscalco, and D. B. Irons. 1998. Selec-tion of forage-fish schools by murrelets and Tufted Puffins in Prince William Sound, Alaska. Condor. 100:286-297.  Paine, R.T., J.L. Ruesink, A. Sun, E.L. Sou-lanille, M.J. Wonham, C.D.H. Harley, D.R. Brumbaugh, and D.L. Secord. 1996. Trouble on oiled waters: lessons from the Exxon Valdez oil spill. Ann. Rev. Ecol. Syst. 27:197-235. Palmerini, P., and C. N. Bianchi. 1994. Biomass measurements and weight-to-weight con-version factors: a comparison of methods applied to the mussel Mytilus gallo-provincialis. Marine Biol. 120:273-277. Parker, K. S., T. C. Royer, R. B. Deriso. 1995. High-latitude climate forcing and tidal 112     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      mixing by the 18.6-year lunar nodal cycle and low-frequency recruitment trends in Pacific halibut (Hippoglossus stenolepis). p. 447-459 In: R. J. Beamish, (ed.). Cli-mate Change and Northern Fish Popula-tions. National Research Council of Can-ada, Ottawa, ON, Canada 121.  Parker, K. S., T. C. Royer, R. B. Deriso. 1995. High-latitude climate forcing and tidal mixing by the 18.6-year lunar nodal cycle and low-frequency recruitment trends in Pacific halibut (Hippoglossus stenolepis). p. 447-459 In: R. J. Beamish, (ed.). Cli-mate Change and Northern Fish Popula-tions. National Research Council of Can-ada, Ottawa, ON, Canada 121.  Parker, R.R. 1968. Marine mortality schedules of pink salmon of the Bella Coola River, Central British Columbia. J. Fish. Res. Board Canada. 25:757-794. Parks, N.B. and H. Zenger. 1979.  Trawl survey of demersal fish and shellfish resources in Prince William Sound, spring 1978. NWAFC processed report 79-2, North-west and Alaska Fisheries Center, Seattle, WA, 49 p.  Parrish, J. K. 1995. Influence of group size and habitat type on reproductive success in Common Murres (Uria aalge). Auk. 112:390-401. Parsons, T. R., M. Takahashi, and B. Hargrave. 1988. Biological Oceanographic Proc-esses, 3rd Edition. Pergamon Press, New York. Patten, S. M. 1994. Assessment of injury to sea ducks from hydrocarbon uptake in Prince William Sound and the Kodiak Archipel-ago, Alaska, following the Exxon Valdez oil spill. NRDA Bird Study Number 11 Final Report. Paul, A.J., J.M. Paul, and R.L. Smith.   1988.  Energy ingestion and conversion rate in Theragra chalcogramma.  Proceedings of the International Symposium on Biologi-cal Management of Walleye Pollock, Nov. 1988, Anchorage, Alaska. Pauly, D. (ed.). 1998. Use of Ecopath with Eco-sim to evaluate Strategies for Sustainable Exploitation of Multi-Species Resources: Proceedings of a Workshop held at the Fisheries Centre, University of British Columbia, Vancouver, BC, Canada, March 25-27, 1998. Fisheries Centre Re-search Reports 6(2), 49 p. Pauly, D. 1984. Fish Population Dynamics in Tropical Waters: a manual for use with programmable calculators. ICLARM Studies and Reviews 8. 325 p. Pauly, D. and V. Christensen (eds.). 1996. Mass-Balance Models of North-eastern Pacific Ecosystems: Proceedings of a Workshop held at the Fisheries Centre, University of British Columbia, Vancouver, B.C., Can-ada. Fisheries Centre Research Reports 4(1), 132 p. Pauly, D. and V. Christensen. 1993. Stratified models of large marine ecosystem: a gen-eral approach and an application to the South China Sea, p. 148-174. In: K. Sherman, L.M. Alexander and B.D. Gold (eds.) Large marine ecosystems: stress, mitigation and sustainability. AAAS Symposium. AAAS Press, Washington, D.C. Pauly, D., J. Dalsgaard, T. A. Okey, R. Powell, and S. Pimm. 1998. Mass-balance food web Ecosystem models as an alternative approach for combining multiple informa-tion sources in fisheries. p. 815 In: F. Funk, T. J. Quinn II, J. Heifetz, J. N. Ian-elli, J. E. Powers, J. F. Schweigert, P. J. Sullivan, C. I. Zhang (Eds.) Fishery Stock Assessment Models, University of Alaska Sea Grant, Fairbanks. Pauly, D., M. L. Soriano-Bartz and M. L. D. Palomares. 1993. Improved construction, parameterization and interpretation of steady-state ecosystem models, p. 1-13. In V. Christensen and D. Pauly (eds.) Tro-phic models of aquatic ecosystems. ICLARM Conf. Proc. 26, 390 p. Pauly, D., T. Pitcher, and D. Preikshot (eds.) 1998. Back to the Future: Reconstructing the Strait of Georgia Ecosystem Fisheries Centre Research Reports, Vol. 6(5). Uni-versity of British Columbia Fisheries Centre, Vancouver. Piatt, J. F., D. D. Roby, L. Henkel, and K. Neu-man. 1998. Habitat use, diet and breeding biology of Tufted Puffins in Prince Wil-liam Sound, Alaska. Northwest Nat (In press). Pitcher, K. W., and D. G. Calkins. 1979. Biology of the harbor seal, Phoca vitulina rich-ardsi, in the Gulf of Alaska. U. S. Dept. Commerce/NOAA/OCSEAP, Environ. Assess. Alaskan Cont. Shelf Fin. Rep. Principal Invest. 19(1983):231-310. ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     113 Polovina, J. 1996. Sharks, p. 23-24. In: D. Pauly and V. Christensen (eds.). Mass-balance models of North-eastern Pacific Ecosys-tems. Fisheries Centre Research Report 4(1). Polovina, J. 1996. Spiny Dogfish, p. 45. In: D. Pauly and V. Christensen (eds.). Mass-balance models of North-eastern Pacific Ecosystems. Fisheries Centre Research Report4(1). Powell, R. D. and S. L. Pimm. 1999. Toward a mass-balanced dynamic model of Prince William Sound ecosystem for the post-spill period 1994-1996. Restoration Pro-ject Annual Report 98330b submitted to the Exxon Valdez Oil Spill Trustee Coun-cil, Anchorage.  Pullum, G.K. 1991. The great Eskimo vocabu-lary hoax. P. 159-171. In Pullum, G.K. The great Eskimo vocabulary hoax and other irreverant essays on the study of language. Chicago. The University of Chicago Press, 236 p. Pullum, G.K. and W.A. Ladusaw. 1986. Phonetic symbol guide. Chicago, The University of Chicago Press, 266 p. PWSFERPG 1993. Sound Ecosystem Assess-ment: Initial Science Plan and Monitoring Program. Prince William Sound Research Planning Group. Report Number 1. 24, November 1993. PWSFERPG, Cordova, Alaska, 111 p. Raymont, J. E. G. 1983. Plankton and productiv-ity in the oceans: Volume 2: zooplankton. Pergamon Press. New York. Reid, W. V. 1987. Life-history evolution in the Glaucous-winged Gull. Ph.D. Thesis, Univ. of Washington, Seattle, WA. Ricker, W. E. 1975. Computation and interpreta-tion of biological statistics of fish popula-tions. Bull. Fish. Res. Board Can. 191:382 p. Robertson, I. 1974. The food of nesting Double-crested and Pelagic Cormorants at Man-darte Island, British Columbia, with notes on feeding ecology. Condor. 76:346-348. Robinson, C. L. and D. M. Ware. 1994. Model-ling pelagic fish and  plankton trophody-namics off Southwestern Vancouver Is-land, British Columbia. Can. J. Fish. Aquat. Sci. 51: 1737-1751. Rosenthal, R. J. 1983. Shallow water fish assem-blages in the northeastern Gulf of Alaska: habitat evaluation, species composition, abundance, spatial distribution and tro-phic interaction, p. 451-540. In: Environ-mental assessment of the Alaskan conti-nental shelf. Final reports of principal in-vestigators, Vol. 17, biological studies. National Oceanic and Atmospheric Ad-ministration/National Ocean Service, Of-fice of Oceanography and Marine Ser-vices, Anchorage, Alaska. Rosenthal, R. J., D. C. Lees, and T. M. Rosen-thal. 1977. Ecological assessment of sub-littoral plant communities in northern Gulf of Alaska. Final report to National Marine Fisheries Service, Auke Bay Laboratory, Juneau, Alaska. Rosenthal, R. J., V. Moran-O’Connell, and M. C. Murphy. 1988. Feeding ecology of ten species of rockfishes (Scorpaenidae) from the Gulf of Alaska. Calif. Fish and Game 74(1):16-37. Rosenthal, R. J., W. D. Clarke, and P. K. Dayton. 1974. Ecology and natural history of a stand of giant kelp, Macrocystis pyrifera, off Del Mar, California. Fishery Bulletin 72:670-684. Sanger, G. A. 1987. Trophic levels and trophic relationships of seabirds in the Gulf of Alaska. p. 229-257. In:: J. P. Croxall (ed.). Seabirds: feeding ecology and role in marine ecosystems. Cambridge Univer-sity Press, New York. Sanger, G. A., and R. D. Jones, Jr. 1982. Winter feeding ecology and trophic relationships of oldsquaws and white-winged scoters on Kachemak Bay, Alaska, p. 20-28. In: Nettleship, D. N., G. A. Sanger, and P. F. Springer (eds.). Marine birds: their feed-ing ecology and commercial fisheries re-lationships. Proceedings of the Pacific Seabird Group Symposium, Can. Wildl. Serv. Special Publication. Sano, O.  1960. The investigation of salmon shark as a predator on salmon in the North Pacific, 1959. Bull. Hokkaido Reg. Fish. Res. Lab., No. 22: 68-82, 2 pls. (In Japanese with English abstract) Sano, O.  1962. The investigation of salmon shark as a predator on salmon in the North Pacific, 1960. Bull. Hokkaido Reg. Fish. Res. Lab., No. 24:148-162.  (In Japanese with English abstract.) Saulitis, E.L., C.O. Matkin, K. Heise, L. Barrett Lennard, and G.M. Ellis. 1998. Foraging strategies of sympatric killer whale (Or-cinus orca) population in Prince William Sound, Alaska. Marine Mammal Science, In press.  114     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Saunders, M.W., and G.A. McFarlane. 1993. Age and length at maturity of the female spiny dogfish, Squalus acanthias, in the Strait of Georgia, British Columbia, Can-ada. Environm. Biol. of Fishes 38: 49-57. Seto, N. W. H., and S. Conant. 1996. The effects of rat (Rattus rattus) predation on the re-productive success of the Bonin Petrel (Pterodroma hypoleuca) on Midway At-oll. Col. Waterbirds. 19:171-185. Smith, K. L. and R. S. Kaufmann. 1999. Long-term discrepancy between food supply and demand in the deep eastern north Pa-cific. Science 284:1174-1177. Smith, R.L., A.J. Paul and J.M. Paul.  1991.  Daily ration estimates for yellowfin sole, Limanda aspera (Pallas), based on labora-tory consumption and growth.  J. Fish Biol. 38: 343-250. Smith, S. E., D. W. Au, and C. Show. 1998. In-trinsic rebound potentials of 26 species of Pacific sharks. Marine and Freshwater Research 49(7):663-678. Spies, R. B., S. D. Rice, D. A. Wolfe, B. A. Wright. 1996. The effects of the Exxon Valdez oil spill on the Alaskan coastal en-vironment. p. 1-16. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (eds.). Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18, Bethesda. Stalmaster, M. V., and J. A. Gessaman. 1984. Ecological energetics and foraging behav-ior of overwintering Bald Eagles. Ecol. Monographs. 54:407-428. Stillwell, C.E. and N.E. Kohler. 1982. Food, feeding habits and estimates of daily ra-tion of the shortfin mako (Isurus oxy-rinchus) in the northwest Atlantic. Can. J. Fish. Aquat. Sci. 1975, 352p. Stocker, M. (Editor). 1994. Groundfish stock assessments for the West-coast of Canada in 1993 and recommended yield options for 1994. Can. Tech. Rep. Fish. Aquat. Sci. 1975, 352p.  Sundberg, K., L. Deysher, L. McDonald. 1996. Intertidal and supratidal site selection us-ing a geographical information system, p. 167-176. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (eds.). Pro-ceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18, Bethesda. Tanaka, S. 1980. Biological investigation of Lamna ditropis in the north-western wa-ters of the North Pacific. In: Report of in-vestigation on sharks as a new marine re-source (1979). Published by: Japan Ma-rine Fishery Resource Research Center, Tokyo. Tanaka, S. 1986. Sharks. Iden (Heredity), 40:19-22 Taylor, F.H.C. 1967. Midwater trawl catches from Queen Charlotte Sound and the open ocean adjacent to the Queen Charlotte Is-lands. Fish. Res. Board Can. Tech. Rep. 11: 1-44. Thomas, G.L. and T.B. Stables.  1995.  Winter 1995 estimate of the prespawning bioi-mass of walleye pollock in Prince Wil-liam Sound, Alaska. Final report to the Alaska Dept. of Fish and Game, Prince William Sound Science Center, Cordova, Alaska. Tranter, D. J. 1976. Herbivore production, p. 186-224. In: D. H. Cushing and J. J. Walsh (eds.). The Ecology of the Seas. W. B. Saunders, Philadelphia. Valiela, I. 1955. Marine Ecological Processes, 2nd Edition. Springer-Verlag, New York. Van Pelt, T. I., J. F. Piatt, B.K. Lance, and D.D. Roby. 1997. Proximate composition and energy density of some North Pacific for-age fishes. Comp. Biochem. Physiol. Vol 118A:4:1393-1398. Venier, J. 1996a. Balancing the Straight of Georgia Model, p. 74-80. In: D. Pauly and V. Christensen (eds.). Mass-balance mod-els of North-eastern Pacific Ecosystems. Fisheries Centre Research Report 4(1), 132 p. . Venier, J. 1996b. Pacific Halibut, p. 43-45. In: D. Pauly and V. Christensen (eds.). Mass-balance models of North-eastern Pacific Ecosystems. Fisheries Centre Research Report 4(1). Vermeer, K. 1981. Foods and populations of surf scoters in British Columbia. Wildfowl 32:107-116. Vermeer, K. 1982. Food and distribution of three Bucephala species in British Columbia waters. Wildfowl 33:22-30. Vermeer, K., and N. Bourne. 1982. The white-winged scoter diet in British Columbian waters: resource partitioning with other scoters, p. 30-38. In: Nettleship, D. N., G. A. Sanger, and P. F. Springer (eds.). Ma-rine birds: their feeding ecology and commercial fisheries relationships. Pro-ceedings of the Pacific Seabird Group ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     115Symposium, Can. Wildl. Serv. Special Pub. Vetter, E.W. 1994. Hotspots of benthic produc-tion. Nature 327:47. Vetter, E.W. 1995. Detritus-based patches of high secondary production in the near-shore benthos. Marine Ecology Progress Series 120:251-262. Von Ziegesar, O., E. Miller and M. E. Dahlheim, 1994, Impacts on Humpback whales in Prince William Sound. In: T. R. Loughlin, (ed.). Marine Mammals and the Exon Valdez. Academic Press Inc., USA. Waite, J. M., M. E. Dahlheim, R. C. Hobbs, S. A. Mizroch, O. von Ziegesar-Matkin, J. M. Straley, L. M. Herman, and J. Jacob-sen. 1999.  Evidence of a feeding aggre-gation of  humpback whales (Megaptera novaeangliae) around Kodiak Island, Alaska. Marine Mammal Science, 15(1):210-220. Walker, R.V., K.W. Meyers, N.D. Davis, H.R. Carlson, and K.D. Friedland. 1999. U.S. releases and recoveries of salmonid data storage tags and disk tags in the North Pacific Ocean and Bering Sea, 1999.  (NPAFC Doc. 412.)  FRI-UW-9910.  Fisheries Research Institute, University of Washington, Seattle WA.  20p. Walters, C. 1998. Ecosim and Ecospace: basic considerations, p. 11-14. In:: Pauly, D. (ed.). Use of Ecopath with Ecosim to Evaluate Strategies for Sustainable Ex-ploitation of Multi-Species Resources, Fisheries Centre Research Report 6(2), 49 p.  Walters, C., D. Pauly, and V. Christensen. (In press). Ecospace: prediction of mesoscale spatial patterns in trophic relationships of exploited ecosystems, with emphasis on the impacts of marine protected areas. In press in Ecosystems Wespestad, V.G. and S.M. Fried. 1983. Review of the Biology and Abundance Trends of Pa-cific Herring (Clupea harengus pallasii), p. 17-29. In W.S. Wooster (ed.). From Year to Year: Interannual Variability of the Envi-ronment and Fisheries of the Gulf of Alaska and the Eastern Bering Sea. Washington Sea Grant Publication. University of Washington, Seattle. Wetzel, R. G. 1983. Limnology, 2nd Edition. Saunders College Publishing, New York, 765 p. Wheelwright, J. 1994. Degrees of Disaster: Prince William Sound: How Nature Reels and Rebounds. Simon and Schuster, New York, 348 p. Wilderbuer, T.K. and E.S. Brown. 1993. Flatfish, p. 3, 1-3, 21. In: Groundfish planning team (eds.). Stock Assessment and Fish-ery Evaluation Report for the Groundfish Resources of the Gulf of Alaska as Pro-ject for 1994. North Pacific Fishery Man-agement Council, Anchorage, Alaska. Willette, T.M., E. Debevec. J. Johnson. 1995a. Sound ecosystem assessment: salmon predation. Exxon Valdez Oil Spill Resto-ration Project Annual Report (Restoration Project 94320E), Alaska Department of Fish and Game, Anchorage, Alaska. Willette, T.M., E. Debevec. J. Johnson. 1996. Sound ecosystem assessment: salmon predation. Exxon Valdez Oil Spill Resto-ration Project Annual Report (Restoration Project 95320E), Alaska Department of Fish and Game, Anchorage, Alaska. Willette, T.M., G. Carpenter, P. Sadler, M. Pow-ell. 1996. Sound Ecosystem Assessment: salmon growth and mortality. Exxon Val-dez oil spill restoration project annual re-port (Restoration project 95320A), Alaska Department of Fish and Game, Cordova, Alaska. Willette, T.M., M. Clapsadl, P. Saddler, and M. Powell. 1997. Sound ecosystem assess-ment: salmon predation. Exxon Valdez Oil Spill Restoration Project Annual Report (Restoration Project 96320E), Alaska De-partment of Fish and Game, Anchorage, Alaska. Willette, T.M., M.V. Sturdevant, S.C. Jewett, and E. Debevec.  1995b.  Forage fish in-fluence on recovery of injured species: forage fish diet overlap. Exxon Valdez Oil Spill Restoration Project Annual Report (Restoration Project 94163), Alaska De-partment of Fish and Game, Anchorage, Alaska. Wootton, J. T. 1994. Predicting direct and indi-rect effects: An integrated approach using experiments and path analysis. Ecology 75(1):151-165. Wynne, K. 1992. Guide to Marine Mammals of Alaska. Alaska Sea Grant College Pro-gram. University of Alaska Fairbanks. Yang, M.S.  1993.  Food habits of the commer-cially important groundfishes in the Gulf of Alaska in 1990.  NOAA Technical Memorandum NMFS-AFSC-22, 150 p. 116     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Appendices  Appendix 1 List of Contributors to PWS Ecopath Model  Jennifer Allen [3] Prince William Sound Sci. Center 300 Breakwater Avenue P. O. Box 705 Cordova, AK 99574, USA phone: (907) 424-5800 fax: (907) 424-5800 jrallen@grizzly.pwssc.gen.ak.us jrallen@alaska.net  Paul Anderson [1, 4] NMFS Kodiak, AK Phone: (907) 487-5961 paul.j.anderson@noaa.gov  Bill Bechtol [4] Alaska Department of Fish and Game Division of Commercial Fisheries 3298 Douglas Place Homer, Alaska 99603-8027 phone: 907.235-8191 billb@fishgame.state.ak.us  Mary Anne Bishop [4] Pac NW research station, USFS c/o CRDI, P.O. Box 1460 Cordova AK  99574 phone: 907.424-7212 fax: 907.424-7214 mbishop@eagle.ptialaska.net  Jim Blackburn [1, 4] Alaska Dept. of Fish and Game Kodiak, AK (907) 486-1863 jblackburn@fishgame.state.ak.us      1) Attended January, 1998 scoping meeting 2) Attended March, 1998 model specification meeting 3) Attended October, 1998 model refinement meeting 4) Collaborated via internet, telephone, or in person  James L. Bodkin [1, 2, 4] Coastal Ecosystems USGS/Biological Resources Division Alaska Biological Science Center 1011 E. Tudor Rd. Anchorage,  Alaska  99577 phone: 907.786-3312 fax: 907.786-3636 James_Bodkin@usgs.gov http://www.absc.usgs.gov  Evelyn D. Brown [4] UAF SFOS IMS Box 757220 Fairbanks, AK 99775-7220 phone:  (907)474-5801 ebrown@ims.uaf.edu  Brian Bugh [4] Alaska Department of Fish and Game Division of Commercial Fisheries 333 Raspberry Road Anchorage, Alaska 99518-1599 phone: 907.267-2123 brian_bue@fishgame.state.ak.us  R. Ted Cooney [2, 3, 4] Institute of Marine Sciences University of Alaska, Fairbanks, Alaska 99775 phone: 907.474-5863 fax: 907.474-7204  cooney@ims.alaska.edu  Johanne Dalsgaard [4] Fisheries Centre University of British Columbia 2204 Main Mall Vancouver, BC V6T 1Z4  CANADA phone: 604-822-2731 jtd@dfu.min.dk   Thomas A. Dean [1, 2, 4] Coastal Resources Associates, Inc. 1185 Park Center Dr.  Suite A Vista, CA.  92083 phone:  (760)727-2004 Fax:  (760) 727-2207 Coastal_Resources@Compuserve.com  ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     117  Jane DeCosimo [4] NPFMC 605 W. 4th Ave., Suite 306 Anchorage, AK 99501 phone: (907) 271-2809 fax: (907) 271-2817  David Duffy Alaska Natural Heritage Program UAA Phone: (907) 257-2784 afdcd1@uaa.alaska.edu  Dan Esler [1, 4] AK Biological Sciences Center USGS, 1011 E. Tudor Rd. Anchorage, AK  99503 phone: 907.786-3485 daniel_esler@usgs.gov  David Eslinger [4] Institute of Marine Science University of Alaska Fairbanks Fairbanks, Alaska phone:  (907) 474-1197 fax:   (907) 474-5863 eslinger@ims.alaska.edu  George E. Esslinger [4] Coastal Ecosystems USGS/Biological Resources Division Alaska Biological Science Center 1011 E. Tudor Rd. Anchorage,  Alaska  99577 http://www.absc.usgs.gov  Robert J. Foy [2, 4] Institute of Marine Science School of Fisheries & Ocean Sci. University of Alaska Fairbanks Fairbanks, Alaska phone:  (907)474-5801 foy@ims.alaska.edu   Kathryn J. Frost [1, 4] Marine Mammals Biologist Alaska Department of Fish & Game 1300 College Road Fairbanks, AK 99701 USA phone:  (907) 459-7214     fax:  (907) 452-6410 e-mail:  kfrost@fishgame.state.ak.us     Joy Geiselman [2, 3, 4] Assistant to the Center Director U.S.G.S. AK Biol. Science Center 1011 E. Tudor Road Anchorage, AK 99503-6199 phone: (907)786-3668 fax: (907)786-3636 email:joy_geiselman@usgs.gov  Tracey Gotthardt [4] Biological Sciences Dept. UAA, 707 A Street Anchorage, AK 99501 phone: 907.257-2788 fax: 907.252-2789 attag@uaa.alaska.edu  Andrew Gunther [1,2,3, 4] Applied Marine Sciences 4749 Bennett Drive, Suite L Livermore, California, USA, 94568 phone: 510.451-7936 fax: 510.451-3631 gunther@amarine.com  William J. Hauser [2, 3, 4] Alaska Dept. of Fish and Game – H & R Division 333 Rasberry Road Anchorage, AK 99518-1599 billh@fishgame.state.ak.us  Roderick Hobbs [2, 4] Small Cetacean and Beluga Tasks Leader National Marine Mammal Laboratory AK Fisheries Science Centr, NOAA, NMFS 7600 Sand Point Way N.E., Bin C15700 Seattle, Washington  98115-0070 phone: (206) 526-6278 fax:  (206) 526-6615 rod.hobbs@noaa.gov  Lee Hulbert [4] NMFS Auke Bay Laboratory 11305 Glacier Hwy Juneau, Alaska 99801-8626 phone:  (907) 789-6056 fax:  (907) 789-6094 e-mail: lee.hulbert@noaa.gov       David Irons [2, 3, 4] 118     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      USFWS 1011 E. Tudor Rd.  Anchorage AK 99503 phone: 907.786-3376 fax: 907.276-6847  David_Irons@mail.fws.gov  Gail Irvine [2, 3, 4] USGS – BRD (907) 786-3512 gail_irvine@usgs.gov  Stephen C. Jewett [2, 4] Institute of Marine Science School of Fisheries & Ocean Sciences University of Alaska Fairbanks Fairbanks, AK 99775 phone:  (907) 474-7841 fax:  (907) 474-7204 Email: jewett@ims.uaf.edu Web: http://www.ims.uaf.edu:8000/  Jay Kirsh [4] PWS Science Center P.O. Box 705, Cordova AK  99574 phone: 907.424-5800 fax: 907.424-5820 kirsch@grizzly.pwssc.gen.ak.us  Thomas C. Kline, Jr. [1,2,3, 4] Prince William Sound Sci. Center 300 Breakwater Avenue P. O. Box 705 Cordova, AK 99574, USA phone: (907) 424-5800 fax: (907) 424-5800 tkline@grizzly.pwssc.gen.ak.us http://www.pwssc.gen.ak.us/  Jeff Leer Alaska Native Language Centre University of Alaska, Fairbanks P.O. Box 757680 Fairbanks AK 216 A E1 phone: 907 474 6587 jleer@mosquitonet.com  Craig Matkin [1, 4] North Gulf Oceanic Society P.O. Box 15244 Homer, AK  99603 phone: 907.235-6590 comatkin@xyz.net  Peter McRoy [4] Institute of Marine Sciences University of Alaska, Fairbanks  Fairbanks, AK  99775 phone: 907.474-7783 ffcpm@aurora.alaska.edu  Scott Meyer [4] Fishery Biologist  Alaska Dept. Fish and Game, 3298    Douglas Place, Homer, AK 99603 phone:  (907) 235-8191 fax:  (907) 235-2448,    scottme@fishgame.state.ak.us  Dan H. Monson [2] Coastal Ecosystems USGS/Biological Resources Division Alaska Biological Science Center 1011 E. Tudor Rd. Anchorage,  Alaska  99577 http://www.absc.usgs.gov  Steve Morestad [4] Alaska Department of Fish & Game Commercial Fisheries, Central Region P.O. Box 669 Cordova AK 99574 phone: 907.424-3212  Phillip R. Mundy [3, 4] Fisheries and Aquatic Sciences 1015 Sher Lane Lake Oswego, OR 97034-1744 phone: 503-699-9856 fax: 503-636-6335 mundy@teleport.com  Karen Murphy [3, 4] Chugach National Forest P.O. Box 129 Girdwood AK 99587 phone: (907) 783-3242 Fax: (907) 783-2094 murphy_karen/r10_chugach_glacier@fs.fed.us  Thomas A. Okey [1, 2, 3, 4] Marine Ecologist, Fisheries Centre University of British Columbia 2204 Main Mall Vancouver, BC V6T 1Z4  CANADA phone: 604-822-1950 fax: 604-822-8934 tokey@fisheries.com http://fisheries.com/members/tomokey.htm  ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     119  William Ostrand [2, 3, 4] US Fish and Wildlife Service 1011 E. Tudor Rd.  Anchorage AK 99503 (907) 786-3849 fax: 907.786-3641 william_ostrand@mail.fws.gov  A.J. Paul [1, 4] Institute of Marine Sciences University of Alaska Fairbanks, AK  99775 phone: (907) 224-5261 ffajp@aurora.uaf.edu  Daniel Pauly [1, 2, 3, 4] UBC Fisheries Centre 2204 Main Mall, Vancouver BC V6T 1Z4 phone: 604.822-1201 fax: 604.822-8934 pauly@fisheries.com  Charles ‘Pete’ Peterson [1, 4] UNC - Chapel Hill Chapel Hill, NC phone: 919.726-6841 cpeters@email.unc.edu  Stuart Pimm [1, 2, 3, 4] U.Tenn, Ecology & Evol. Biology 2415 Lakemoor Drive, Knoxville, TN 37920 phone: 423.974-0978 fax: 423-974-0978 StuartPimm@aol.com  Bob Powell [1, 2, 3, 4] U.Tenn, Ecology & Evol. Biology 569 Dabney Hall  Knoxville TN  37916 phone: 423.974-6186 fax: 423.974-0978 qqi@utk.edu  Dave Preikshot [3, 4] Fisheries Centre University of British Columbia 2204 Main Mall Vancouver, BC V6T 1Z4  CANADA phone: 604-822-0618 fax: 604-822-8934 preikshot@fisheries.com     Jennifer Purcell [1, 4] U. of Maryland, CTR. for Env. Sci. Horn Point Lab P.O. Box 775 Cambridge MD 21613 phone: 410.221-8431 purcell@hpl.umces.edu  Jennifer Ruesink Department of Zoology University of Washington  Seattle, Washington ruesink@zoology.ubc.ca  Stanley Senner [2, 3, 4] EVOS Trustee Council Restoration  645 G Street, Ste 402 Anchorage, AK 99501-3451 phone: (907) 278-8012 fax; (907) 276-7178 stans@oilspill.state.ak.us  Tom Shirley [4] UAF - Juneau Juneau, Alaska phone: 907 465-6449 fftcs@uaf.edu  Curtis Smith [3, 4] Alaska Department of Fish & Game Habitat and Restoration Division 333 Raspberry Road Anchorage, AK  99518-1599 Voice: (907) 267-2295 Fax: (907) 267-2464 curtiss@fishgame.state.ak.us  Robert B. Spies [3, 4] Applied Marine Sciences 4749 Bennett Drive, Suite L Livermore, California, USA, 94568 phone: 510.373-7142 fax: 510.373-7834 spies@amarine.com  Staff of the International Pacific Halibut Commission [4] P.O. Box 95009 Seattle, WA 98145 phone: 206-634-1838  fax: 206-632-2983  http://www.iphc.washington.edu/   Molly V. Sturdevant [4] 120     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Fisheries Research Biologist Auke Bay Laboratory, Alaska Fisheries Science Center, NMFS, NOAA 11305 Glacier Highway Juneau, AK  USA  99801-8626 phone:  (907)789-6041 fax:  (907)789-6094 molly.sturdevant@noaa.gov  Joe Sullivan [3] Alaska Dept. of Fish and Game joes@fishgame.state.ak.us  Rob Suryan [3, 4] USFWS robert_suryan@mail.fws.gov  Lisa Thomas [1, 4] U.S.G.S. AK Biological Sci. Center 1011 E. Tudor Road Anchorage, AK 99503-6199 phone: 907. 786-3685 lisa_m_thomas@usgs.gov  Charlie Trowbridge [4] Alaska Department of Fish and Game Division of Commercial Fisheries 3298 Douglas Place Homer, Alaska 99603-8027 phone: 907. 235-8191  fax: 907. 235-2448  charlie_trowbridge@fishgame.state.ak.us  Bob Trumble [4] International Pac. Halibut Commission Seattle, Washington phone: 206.634-1838 fax: 206.632.2983 bob@iphc.washington.edu  John Wilcock [1, 4] Alaska Department of Fish and Game Division of Commercial Fisheries 401 Railroad Avenue P.O. Box 669 Cordova, Alaska 99574-0669 phone: 907.424-3212 fax: 907.424-3235 johnwi@fishgame.state.ak.us      Mark Willette [2, 4] Alaska Department of Fish and Game Division of Commercial Fisheries 401 Railroad Avenue P.O. Box 669 Cordova, Alaska 99574-0669 phone: 907.424-3214 fax: 907.424-3235 markw@fishgame.state.ak.us  Bruce Wright [1, 2, 3, 4] Chief, OOSDAR NMFS, NOAA, DOC 11305 Glacier Highway Juneau, AK  99801 phone: 907-789-6601 fax: 907-789-6608 bruce.wright@noaa.gov  Kelly Zeiner [3] AK Dept. of Natural Resources  phone: (907) 269-8856 kellyz@dnr.state.ak.us ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     121Appendix 2. Workshop Agendas   Fridays Lunch Meeting Agenda  Preparation of March Food Web Workshop  12:15 - 1 pm, January 30, 1998 Quadrant Room, Hotel Captain Cook Anchorage, Alaska   This is an informal lunch meeting during which data-weak components of a straw-man PWS ecosys-tem will be identified for refinement by EVOS-funded investigators, and during which participants can help plan a collaborative food web workshop to be held in March  Lunch for that day (sandwiches, salads, and drinks) will be served in the room for the participants of the meeting.  1. Introduction of Ecopath project managers and EVOS researchers         (10 min) a.  Stuart Pimm b.  Bob Powell c.  Daniel Pauly d.  Tom Okey e. Name, affiliation, and interest of other EVOS researchers   2. Presentation of UBC/Tennessee near-term project aims (materials)      (15 min) a.  Preliminary Ecopath model of PWS (report) b.  Poster outlining key elements of project  * including list of ecosystem components for which information is needed  c.  Guidelines for describing functional groups in PWS ecosystem  3. Planning the March workshop                                                  (15 min) a.  Suggestions for items to consider b.  Suggestions for meeting participants c.  Other matters  4. Further discussion (guided small groups as desired)  Please contact Tom Okey at UBC with questions about the agenda, or the lunch meeting. Fisheries Centre, 2204 Main Mall, Vancouver B.C. Canada V6T 1Z4 604.822-1950, tokey@fisheries.com 122     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996       Food Web Workshop Agenda  Constructing an ECOPATH model of Prince William Sound  March 2-4, 1998 EVOS Restoration Office 245 G Street, Suite 401, Anchorage  Monday 2 Workshop Day 1 in restoration office conference room 0915 - 0920 Welcome to EVOS office …. Stan Senner 0920 - 1925 EVOS program context--ecosystem synthesis …. Andy Gunther 0925 - 0945 Round-table introductions 0945 - 1000 About this workshop …. Daniel Pauly 1000 - 1015 Coffee break 1015 - 1100 Presentation 1 - Introduction and orientation to ECOPATH modelling-D. P. 1100 - 1140 Presentation 2 - Demonstration of preliminary ECOPATH model of Prince William Sound prior to the EVOS …. Tom Okey 1140 - 1200 Questions and discussion 1200 - 1330 Optional sandwich lunch provided in room, or lunch on your own 1330 - 1430 Workshop Session 1:  Review of period and area to be covered.   Moderator: D. Pauly  1430 - 1700 Workshop Session 2:  Definition of ecosystem components (“boxes”) to be included in the models, assignment of components, and discussion of completeness.  Moderator:  D. Pauly Tuesday 3 Workshop Day 2 0900 - 1000 Presentation 3: From static to dynamic models …. Stuart Pimm 1000 - 1020 Coffee break 1020 - 1200 Workshop Session 3: Participants assemble key parameter estimates (Biomass, con-sumption, etc.) for their group.   1200 - 1330 Optional sandwich lunch provided in room, or lunch on your own 1330 - 1500 Workshop Session 4: Assembling a diet matrix.  Mod.: D. Pauly 1500 - 1520 Coffee break 1520 - 1700 Workshop Session 5: Data entry and balancing model.   Moderator: D. Pauly Wednesday 4 Workshop Day 3 0900 - 1030 Workshop Session 6: Definition of major habitat types and species or group affinities to these habitats.  Moderator: T. Okey 1030 - 1050 Coffee break 1050 - 1200 Session 6 (continued):  Data entry using ECOPATH IV. 1200 - 1330 Optional sandwich lunch provided in room, or lunch on your own 1330 - 1500 Workshop Session 7:Discussion of flow networks and ancillary statistics of balanced models. Moderator: D. Pauly 1500 - 1520  Coffee break  1500 - 1700 Workshop Session 8: Wrapping Up -  (a)  What have we learned, and what have we accomplished? (b)  Future Actions--applied uses/identified weaknesses, Moderator: D. Pauly  ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     123 Ecopath Workshop Agenda  A balanced trophic model of PWS: presentation and refinement  October 5, 1998 EVOS Restoration office conference room 645 G Street, Suite 401, Anchorage  0915 - 1920 EVOS program context--ecosystem synthesis …. Bob Spies 0925 - 0935 Round-table introductions (whole room) 0935 - 0950 About this workshop, about ECOPATH modelling …. Daniel Pauly  0945 - 1015 Presentation of ECOPATH model of Prince William Sound, 1994-1996 …. Tom Okey a)  Process of model construction  b)  Trophic structure and collaboration c)  Balancing the PWS trophic model  1015 - 1030 Coffee break  1030 – 1130 Analysis if the PWS food web … Stuart Pimm 1130 – 1200 “What if” scenarios and spatial simulations … Tom Okey  1200 - 1330 Lunch on your own in downtown Anchorage  1330 - 1430 Questions and discussion about the simulations and their implications, including sug-gested improvements to the approach.  1430 - 1445 Plans for the coming year … Daniel Pauly  1445 - 1500 Comments from the Chief Scientist and Peer Reviewers  Close  1600 Informal demonstrations of Ecopath and Alaska FishBase (if desired)  124     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Workshop Agenda  Ecosystem-based stewardship of PWS living resources  by local communities and students: uses of a food web model   September 27, 1999,    1 - 5 pm EVOS Restoration office conference room 645 G Street, Suite 401, Anchorage    1:05 – 1:15  Welcoming remarks …. Hugh Short and Helen Morris 1:15 – 1:30  Brief round-table introductions (whole room) 1:30 – 1:35  About this workshop … Tom Okey 1:35 – 2:00  ECOPATH modeling and its uses …. Daniel Pauly  2:00 – 2:30 Presentation 1 – The ECOPATH model of Prince William Sound,  …. Tom Okey a)  Building the model  b)  Animals, plants, and energy flow in PWS c)  Description, simulated playing, and virtual experiments  2:30 – 2:45  Coffee break  2:45 – 3:30  Presentation 2 – Managing resources and learning through simulation   …. Daniel Pauly and Tom Okey  3:30 – 4:00  Playing the ‘what if’ game … (whole room)  4:00 – 5:00  Questions and discussion in casual format   ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     125 Ecopath Workshop Agenda  A balanced trophic model of PWS:  applications for ecosystem-based management   September 28, 1999    9 am  Alaska Dept. of Fish and Game Division of Commercial Fisheries 333 Raspberry Road, Anchorage  0910 - 1920 Welcoming remarks and description of fisheries management in PWS  …. Stephen Fried 0920 - 0930 Round-table introductions (whole room) 0930 - 1000 About this workshop: ECOPATH modeling and its applications …. Daniel Pauly  1000 - 1030 Presentation 1 - Presentation of ECOPATH model of Prince William Sound, 1994-1996 …. Tom Okey d)  Process of model construction  e)  Description of food web structure  f)  From description to dynamic modeling to management  1030 - 1045 Coffee break  1045 - 1115 Presentation 1 (continued): d)  Ecosim - dynamic temporal simulation  e)  Ecospace - dynamic spatial simulation  1115 - 1200 Questions and discussion  1200 - 1330 Lunch on your own  1330 - 1430 Presentation 2:  Management applications of the PWS model: complimenting existing tools  … Daniel Pauly and Tom Okey   1430 - 1530 Group discussion about the approach, the PWS model, and potential manage-ment applications  1530 - 1545 Summary comments  … any participants or presenters  1545 - 1550    Closing comments …. Stephen Fried  1550  Informal demonstrations if desired 126     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      126 Appendix 3 Monthly Estimates for PWS Zooplankton Parameters (R.T. Cooney, unpublished data; depth of integration = 300 m; PWS area used = 8800 km2 ) Taxa Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Biomass (g⋅m-3)        Total Zooplankton 0.03613 0.00805 0.045890.238620.431350.60514 0.35001 0.22229 0.09457 0.09573 0.079970.06420Zoops <= 1 mg wet wt 0.01881 0.00700 0.021130.136130.239330.41893 0.29484 0.17179 0.04874 0.06520 0.047910.03062Zoops >1 mg wet wt 0.01747 0.00135 0.024480.102490.192020.18621 0.05517 0.05050 0.04583 0.03053 0.032060.03358Copepoda 0.01716 0.00433 0.028940.182370.347310.35525 0.18937 0.12095 0.05252 0.05809 0.044040.02999Pteropoda 0.00086 0.00014 0.000180.001100.007810.12087 0.06825 0.03485 0.00144 0.00010 0.000840.00158Amphipoda 0.00437 0.00008 0.005390.004750.005410.00677 0.00449 0.00456 0.00464 0.00704 0.007850.00866Larvacea 0.00028 0.00022 0.000340.002340.024840.03391 0.06993 0.03600 0.00208 0.01097 0.005660.00035Euphausiacea 0.00826 0.00055 0.004020.009040.019130.00905 0.00690 0.00687 0.00683 0.00472 0.010350.01597Other 0.00520 0.00274 0.007010.039020.026850.07929 0.01107 0.01907 0.02707 0.01481 0.011240.00766        (t⋅km-2)        Total Zooplankton 10.84 2.42 13.7771.59129.41181.54 105.00 66.69 28.37 28.72 23.9919.26Zoops <= 1 mg wet wt 5.64 2.10 6.3440.8471.80125.68 88.45 51.54 14.62 19.56 14.373.06Zoops >1 mg wet wt 5.24 0.41 7.3430.7557.6155.86 16.55 15.15 13.75 9.16 9.6210.07Copepoda 5.15 1.30 8.6854.71104.19106.58 56.81 36.28 15.76 17.43 13.213.00Pteropoda 0.26 0.04 0.050.332.3436.26 20.48 10.45 0.43 0.03 0.250.16Amphipoda 1.31 0.02 1.621.431.622.03 1.35 1.37 1.39 2.11 2.350.87Larvacea 0.08 0.06 0.100.707.4510.17 20.98 10.80 0.62 3.29 1.700.03Euphausiacea 2.48 0.17 1.212.715.742.72 2.07 2.06 2.05 1.42 3.101.60Other 1.56 0.82 2.1011.718.0623.79 3.32 5.72 8.12 4.44 3.370.77        t in PWS             Total Zooplankton 95370 21252 12115062995711387641597570 924026 586846 249665 252727 211108169488Zoops <= 1 mg wet wt 49656 18475 557733593946318391105980 778383 453534 128684 172125 12648126946Zoops >1 mg wet wt 46113 3567 64617270574506938491581 145638 133315 120991 80610 8463488659Copepoda 45302 11431 76402481457916898937860 499937 319295 138653 153358 11626626391Pteropoda 2270 370 475290420618319097 180180 91991 3802 264 22181390Amphipoda 11525 201 14238125431428817875 11851 12045 12239 18575 207127616Larvacea 742 570 90861786557289528 184618 95049 5481 28963 14938304Euphausiacea 21806 1452 10613238665050323892 18216 18124 18031 12461 2731114054Other 13724 7228 1851410301070884209318 29225 50342 71460 39106 296636740                Continued on next page…                       ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     127P/B Ratios (monthly)        Total Zooplankton 0.40 0.30 0.400.801.502.30 1.60 0.80 0.70 0.40 0.400.40Zoops <= 1 mg wet wt 0.50 0.30 0.501.003.004.50 2.00 1.00 0.70 0.50 0.500.50Zoops >1 mg wet wt 0.10 0.20 0.401.001.500.50 0.40 0.30 0.20 0.20 0.100.10Copepoda 0.30 0.15 0.250.401.201.60 1.40 1.20 0.50 0.40 0.300.30Pteropoda 0.05 0.02 0.100.400.800.40 0.30 0.30 0.20 0.20 0.150.08Amphipoda 0.10 0.10 0.40 0.20 0.20 0.10 0.100.10Larvacea 0.10 0.10 0.200.300.400.40 0.30 0.30 0.30 0.20 0.200.20Euphausiacea 0.05 0.08 0.25 0.20 0.20 0.20 0.120.10Other 0.10 0.05 0.30 0.20 0.10 0.10 0.100.05        Production (t⋅month-1)        Total Zooplankton 38148 6376 4846050396517081463674410 1478442 469476 174765 101091 8444367795Zoops <= 1 mg wet wt 24828 5542 2788635939418955174976912 1556766 453534 90079 86063 6324113473Zoops >1 mg wet wt 4611 713 25847270574760407245791 58255 39994 24198 16122 84638866Copepoda 13591 1715 1910019258311002781500576 699912 383154 69326 61343 348807917Pteropoda 114 7 48116216495127639 54054 27597 760 53 333111Amphipoda 1152 20 1424125428585363 4740 2409 2448 1858 2071762Larvacea 74 57 18218532622935811 55385 28515 1644 5793 298861Euphausiacea 1090 116 10613580126267168 4554 3625 3606 2492 32771405Other 1372 361 2777257532126562795 8767 10068 7146 3911 2966337        Q/B     (Ingest⋅month1 / biomass)       Total Zooplankton 1.33 1.00 1.332.675.007.67 5.33 2.67 2.33 1.33 1.331.33Zoops <= 1 mg wet wt 1.67 1.00 1.673.3310.0015.00 6.67 3.33 2.33 1.67 1.671.67Zoops >1 mg wet wt 0.33 0.67 1.333.335.001.67 1.33 1.00 0.67 0.67 0.330.33Copepoda 1.00 0.50 0.831.334.005.33 4.67 4.00 1.67 1.33 1.001.00Pteropoda 0.17 0.07 0.331.332.671.33 1.00 1.00 0.67 0.67 0.500.27Amphipoda 0.33 0.33 0.330.330.671.00 1.33 0.67 0.67 0.33 0.330.33Larvacea 0.33 0.33 0.671.001.331.33 1.00 1.00 1.00 0.67 0.670.67Euphausiacea 0.17 0.27 0.330.500.831.00 0.83 0.67 0.67 0.67 0.400.33Other 0.33 0.17 0.500.831.001.00 1.00 0.67 0.33 0.33 0.330.17          128     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      128  Appendix 3 (cont.)  Monthly Estimates for PWS Nearshore Zooplankton Parameters (inshore of 20 m isobath) (R. Foy, unpublished data)  Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec mean Biomass (g⋅m-3)              Carnivorous jellies 0.06 0.00 0.01 0.10 0.20 1.11 0.61 0.44 0.36 0.28 0.07 0.12 0.28 Omnivorous zooplankton 0.02 0.00 0.03 0.04 0.05 0.13 0.13 0.10 0.06 0.02 0.01 0.01 0.05 Herbivorous zooplankton 0.07 0.00 0.06 0.09 0.13 0.14 0.06 0.11 0.06 0.01 0.00 0.01 0.06               Biomass (t⋅km-2)              Carnivorous jellies 0.55 0.01 0.06 1.02 1.98 11.13 6.08 4.36 3.59 2.82 0.65 1.21 2.789 Omnivorous zooplankton 0.16 0.04 0.28 0.40 0.52 1.28 1.26 1.01 0.61 0.22 0.12 0.08 0.497 Herbivorous zooplankton 0.66 0.03 0.58 0.93 1.29 1.42 0.57 1.09 0.58 0.06 0.04 0.12 0.614               Biomass (t⋅km-2; PWS-wide)              Carnivorous jellies 0.09 0.00 0.01 0.16 0.31 1.76 0.96 0.69 0.57 0.45 0.10 0.19 0.440 Omnivorous zooplankton 0.025 0.006 0.044 0.063 0.082 0.204 0.199 0.160 0.097 0.034 0.019 0.012 0.079 Herbivorous zooplankton 0.105 0.005 0.092 0.148 0.204 0.225 0.090 0.174 0.092 0.010 0.006 0.019 0.097               total mt (0-20m stratum)              Carnivorous jellies 780 21 79 1446 2813 15780 8617 6187 5093 3999 927 1710 3954.29 Omnivorous zooplankton 227 52 395 563 731 1820 1783 1431 868 305 170 109 704.49 Herbivorous zooplankton 935 41 826 1325 1824 2013 803 1552 820 88 57 171 871.21               P/B ratio (year-1)              Carnivorous jellies 0.15 0.04 0.04 0.28 0.52 2.98 1.46 1.12 0.94 0.77 0.21 0.33 8.82 Omnivorous zooplankton 0.71 0.07 0.72 1.01 1.30 1.50 0.58 1.17 0.62 0.07 0.02 0.13 7.90 Herbivorous zooplankton 0.66 0.30 2.90 2.45 2.01 5.19 4.96 4.15 2.52 0.89 0.64 0.32 27.00               Q/B ratio (year-1)              Carnivorous jellies 0.50 0.13 0.13 0.93 1.72 9.94 4.85 3.73 3.15 2.56 0.69 1.09 29.41 Omnivorous zooplankton 2.37 0.23 2.39 3.36 4.34 5.01 1.93 3.90 2.06 0.22 0.08 0.43 26.33 Herbivorous zooplankton 2.22 1.01 9.67 8.18 6.69 17.29 16.54 13.82 8.40 2.98 2.14 1.06 90.00  ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     129 Appendix 4 Derivation of diet compositions of forage fish, 1994-1996  (data are from APEX-SEA program, provided by M. Sturdevant) Tables A4-1 to A4-7 (inside the following boxes) show proportions and percentages of prey cate-gories (defined for the PWS model) for seven species of forage fishes. Diets were derived from the APEX-SEA project data contained in each box (data provided by M. Sturdevant).  Table A4-8 shows seasonal changes in the diet composition of juvenile Pacific herring (R. J. Foy, UAF Institute of Marine Sciences, unpublished data). Box A4-1. Pacific Herring diet data Prey May-94 Jun-94 Jul-94 Aug-94 Sep-94 Nov-94 Jul-95 Oct-95 Jul-96 Mean Large calanoids 48.8 37.1 36.5 75.4 14.3 6.2 8.4 47.7 33.3 34.2 Small calanoids 0.2 10.1 4.1 2.8 29.3 17.5 80.2 8.2 24.1 19.6 Larvaceans 0.0 1.1 2.8 1.1 5.2 7.8 1.5 16.1 11.5 5.2 Cladocerans 0.0 0.2 0.0 0.0 0.1 0.0 0.9 0.0 0.2 0.2 Malacostracans 44.8 25.0 0.7 2.9 19.3 0.2 0.3 0.8 3.0 10.8 Euphausiids 0.0 1.8 13.1 0.0 13.3 51.1 0.1 13.3 0.5 10.4 Hyperiid amphipods 0.9 3.4 6.1 13.8 7.5 4.7 0.6 8.6 5.7 5.7 Zoeae 0.1 4.3 1.6 2.9 1.2 0.0 0.0 0.0 0.0 1.1 Chaetognaths 0.0 0.0 0.0 0.0 0.0 0.3 0.0 0.4 0.1 0.1 Fish 4.5 10.3 19.9 0.5 5.1 0.0 0.0 0.0 0.1 4.5 Decapods 0.6 0.3 2.6 0.0 2.6 0.0 1.6 0.6 20.3 3.2 Gastropods 0.0 4.5 9.5 0.0 0.6 2.7 1.0 0.3 0.6 2.1 Invertebrate eggs 0.0 1.5 0.1 0.4 1.4 0.0 0.0 0.0 0.0 0.4 Barnacles 0.1 0.2 1.1 0.0 0.1 0.0 0.1 0.0 0.3 0.2 Others 0.0 0.2 0.1 0.0 0.1 9.4 5.4 3.9 0.4 2.2 Polychaetes 0.0 0.0 1.6 0.0 0.0 0.0 0.0 0.0 0.0 0.2         Table A4-1. Derivation of Pacific Herring diet based on above data from APEX-SEA program.      Prey category proportion % in diet      Herbivorous Zooplankton 0.592 59.2      Omnivorous Zooplankton 0.326 32.6      Shallow Sm. Epifauna 0.082 8.2       Box A4-2. Capelin diet data Prey May-94 Jun-94 Jul-94 Jul-95 Oct-95 Mean Euphausiids 1.3 5.6 55.6 0.0 92.6 31.0 Fish 60.3 24.7 0.0 0.0 0.0 17.0 Hyperiid amphipods 17.5 1.0 0.0 0.0 0.0 3.7 Malacostracans 4.4 3.8 0.1 0.0 0.0 1.7 Zoeae 0.6 3.7 3.8 0.0 0.0 1.6 Small calanoids 0.2 5.0 0.0 88.6 7.0 20.2 Large calanoids 13.8 55.8 0.0 0.0 0.0 13.9 Larvaceans 0.0 0.0 37.2 0.0 0.0 7.5 Other 0.0 0.1 3.3 4.6 0.4 1.7 Gastropods 1.3 0.2 0.0 6.7 0.0 1.6 Gammarid amphipods 0.4 0.0 0.0 0.0 0.0 0.1 Decapods 0.3 0.0 0.0 0.0 0.0 0.1     Table A4-2. Derivation of capelin diet based on above data from APEX-SEA program.   Prey category proportion % in diet   Omnivorous Zooplankton 0.5 55.0   Herbivorous Zooplankton 0.4 41.6   Shallow Sm. Epifauna 0.0 3.4    130     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Box A4-3. Sandlance diet data Prey May-94 Jun-94 Jul-94 Sep-94 Jul-95 Jul-96 mean Small calanoids 11.5 12.7 60.5 23.0 62.7 66.7 39.5 Large calanoids 82.0 6.0 16.4 32.1 10.8 12.2 26.6 Larvaceans 0.0 2.1 6.7 2.4 18.3 8.5 6.3 Cladocerans 0.0 0.6 0.3 0.0 0.3 0.5 0.3 Fish 2.5 73.2 2.2 0.0 0.0 0.7 13.1 Malacostracans 2.6 1.8 1.0 27.2 0.0 0.0 5.4 Euphausiids 0.5 0.5 0.5 5.0 0.1 0.1 1.1 Zoeae 0.1 0.2 0.6 3.7 0.0 0.0 0.8 Hyperiid amphipods 0.0 0.3 2.2 0.6 0.5 0.2 0.6 Barnacles 0.0 0.3 1.8 0.2 1.3 5.8 1.6 Invertebrate eggs 0.7 1.2 5.4 1.4 0.0 0.0 1.5 Others 0.0 0.1 0.1 0.0 4.9 2.0 1.2 Decapods 0.0 0.4 0.1 2.4 0.3 2.5 1.0 Gastropods 0.0 0.4 1.4 1.4 0.7 0.8 0.8 Polychaetes 0.0 0.1 0.6 0.0 0.0 0.0 0.1 Bivalves 0.0 0.0 0.0 0.7 0.0 0.0 0.1     Table A4-3. Derivation of sandlance diet based on above data from APEX-SEA program.   Prey category Proportion % in diet   Herbivorous Zooplankton 0.727 72.7   Omnivorous Zooplankton 0.210 21.0   Shallow Sm. Epifauna 0.062 6.2    Box A4-4. Pink Salmon fry diet data Prey May-94 Jun-94 Jul-94 Aug-94 Sep-94 Jul-95 Jul-96 Mean Fish 24.0 28.2 32.4 35.6 34.3 22.2 79.1 36.5Large calanoids 36.1 6.9 11.8 31.1 42.7 13.3 2.7 20.7Small calanoids 7.9 17.1 10.0 0.0 0.0 0.6 0.2 5.1Larvaceans 0.1 5.4 8.3 3.7 3.1 1.0 6.5 4.0Cladocerans 0.0 3.2 0.1 0.0 0.0 0.4 0.0 0.5Gastropods 1.1 21.2 28.3 0.1 0.0 33.7 0.6 12.1Decapods 0.0 0.4 1.0 17.1 1.8 1.3 4.2 3.7Harpact. Copepods 0.9 0.3 0.1 0.0 0.0 0.0 0.0 0.2Polychaetes 0.8 0.2 0.1 0.0 0.0 0.0 0.0 0.2Gammarid amphipods 0.6 0.3 0.1 0.0 0.7 0.0 0.0 0.2Euphausiids 23.1 9.8 1.1 3.5 4.2 14.9 0.2 8.1Hyperiid amphipods 1.1 0.5 3.2 7.9 11.0 7.9 3.7 5.0Barnacles 1.6 1.3 0.5 0.1 0.3 0.1 0.6 0.6Malacostracans 0.0 1.7 0.3 0.6 0.9 0.1 0.8 0.6Chaetognaths 0.0 0.0 0.0 0.0 0.0 4.2 0.2 0.6Zoeae 0.8 0.9 1.3 0.2 0.7 0.0 0.0 0.6Invertebrate eggs 0.3 0.6 0.5 0.0 0.0 0.0 0.0 0.2Others 1.7 1.8 1.0 0.1 0.2 0.1 0.2 0.7Insects 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.1     Table A4-4. Derivation of pink salmon fry diet based on above data from APEX-SEA program.     Prey category Proportion % in diet     Fish 0.365 36.5     Herbivorous zooplankton 0.303 30.3     Carnivorous zooplankton 0.156 15.6     Shal. Sm. Epibenthos 0.164 16.4      ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     131 Box A4-5. Chum salmon fry diet data Prey May-94 Jun-94 Jul-94 Aug-94 Sep-94 Jul-95 Jul-96 MeanFish 83.1 72.2 86.5 3.5 37.6 29.6 12.1 46.4Large calanoids 14.1 5.6 0.3 69.1 12.9 3.8 0.0 15.1Larvaceans 0.4 0.6 2.2 11.0 8.6 5.9 5.4 4.9Small calanoids 0.1 1.6 0.0 0.3 0.0 0.3 0.1 0.3Cladocerans 0.0 1.4 0.0 0.1 0.0 0.0 0.0 0.2Decapods 0.0 0.9 0.0 0.4 2.1 2.5 24.8 4.4Gastropods 0.0 8.8 11.0 0.1 0.1 0.1 0.9 3.0Barnacles 0.0 1.5 0.0 6.6 0.1 0.0 12.1 2.9Gammarid amphipods 0.3 0.2 0.0 0.0 1.2 0.0 0.0 0.2Polychaetes 0.1 0.5 0.0 0.0 0.0 0.0 0.0 0.1Hyperiid amphipods 0.0 0.4 0.0 2.1 8.4 41.3 41.7 13.4Malacostracans 0.0 0.5 0.0 0.0 17.9 0.0 1.0 2.8Chaetegnaths 0.0 0.0 0.0 0.0 0.0 16.4 0.9 2.5Gelatenous zooplankton 0.0 0.0 0.0 6.3 2.7 0.0 0.0 1.3Euphausiids 0.1 1.2 0.0 0.0 7.2 0.1 0.0 1.2Others 1.7 2.2 0.0 0.0 0.1 0.0 0.6 0.7Zoeae 0.0 2.0 0.0 0.3 0.1 0.0 0.0 0.3Invertebrate eggs 0.0 0.1 0.0 0.0 1.0 0.0 0.0 0.2Insect 0.0 0.0 0.0 0.0 0.0 0.0 0.5 0.1    Table A4-5. Derivation of chum salmon fry diet based on above data from APEX-SEA program.    Prey category Proportion % in diet    Fish 0.464 46.4    Herbivorous zooplankton 0.206 20.6    Carnivorous zooplankton 0.222 22.2    Shallow Sm. Epibenthos 0.106 10.6     Box A4-6. Eulachon diet data Prey Nov-94 Oct-95 mean Euphausiids 46.9 97.8 72.3 Malacostracans 53.1 1.1 27.1 Small calanoids 0.0 0.6 0.3 Gastropods 0.0 0.5 0.2    Table A4-6. Derivation of eulachon diet based on above data from APEX-SEA program.  Prey category Proportion % in diet  Omnivorous Zooplankton 0.994 99.4  Herbivorous Zooplankton 0.003 0.3  Shallow Sm. Epifauna 0.002 0.2     132     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      Box A4-7. Small Pacific cod diet data Prey Category May-94 Jun-94 Jul-94 Sep-94 meanLarge calanoids  97.9 32.0 18.5 0.2 37.2Small calanoids 1.9 16.6 10.5 0.4 7.3Larvaceans 0.0 0.7 1.5 0.0 0.6Cladocerans 0.0 1.8 0.1 0.0 0.5Euphausiids 0.0 1.5 9.4 35.0 11.4Others 0.0 0.7 2.6 28.5 8.0Malacostracans 0.0 2.7 2.1 23.1 7.0Hyperiid amphipods 0.0 0.6 4.1 0.6 1.3Zoeae 0.1 1.7 1.8 1.6 1.3Gastropods 0.0 5.1 30.2 0.0 8.8Gammarid amphipods 0.0 12.7 4.7 3.1 5.1Decapods 0.0 2.9 7.2 1.7 3.0Harpacticoid copepods 0.0 3.1 1.2 2.2 1.6Barnacles 0.0 0.9 0.6 0.0 0.4Polychaetes 0.0 0.5 0.0 0.2 0.2Invertebrate eggs 0.1 0.7 0.4 0.0 0.3Fish 0.0 15.7 5.1 3.4 6.1   Table A4-7. Derivation of Pacific cod diet based on above data from APEX-SEA programa.  Prey category Proportion % in diet  Herbivorous Zooplankton 0.455 45.5  Carnivorous Zooplankton 0.290 29.0  Shal. Sm. Epibenthos 0.194 19.4  Small Pelagic Fishes 0.061 6.1  a) Not used in model because data applied only to small individuals; data from Table 2-1 in Yang (1993) were used instead.    Juvenile Pacific Herring Table A4-8. Seasonal dietary changes in juvenile Pacific herring (Robert J. Foy, UAF Institute of Marine Sciences, unpublished data). Prey category Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec mean omnivorous zooplankton 0.75 0.14 0.21 0.20 0.04 0.27 0.37 0.50 0.60 0.56 0.61 0.78 0.42 herbivorous zooplankton 0.25 0.85 0.77 0.80 0.95 0.57 0.60 0.50 0.40 0.44 0.39 0.22 0.56 fish eggs 0.00 0.01 0.02 0.00 0.01 0.16 0.03 0.00 0.00 0.00 0.00 0.00 0.02    ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     133Appendix 5. Input diet compositions (% weight) of animals in Prince William Sound, Alaska, from 1994-1996.  Predator Prey     1      2      3      4      5      6      7      8      9    10    11    12    13    14    15    16    17    18    19    20      1 Transient Orca          -          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -     2 Resident Orca          -          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -     3 Sharks          -          -      0.3          -      5.0          -     3.8          -         -          -          -          -          -          -         -          -         -          -         -          -     4 Halibut      0.5      0.5      2.5          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -     5 Porpoise   59.0          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -     6 Pinnipeds   38.0          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -     7 Lingcod          -          -          -          -          -      0.1      0.5          -         -          -          -          -          -          -     1.0          -         -          -         -          -     8 Sablefish      0.5      0.5      1.0          -  10.0      1.0          -         -         -          -          -          -          -          -         -          -         -          -         -          -     9 Adult Atooth          -          -   15.0          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -   10 Adult Salmon      1.5   74.0   32.3   13.3          -   14.0   13.6          -         -          -          -          -     8.0          -         -          -         -          -         -          -   11 Pac. Cod          -      0.5      3.5      1.2  14.0      1.0          -    0.7          -          -          -          -          -          -         -          -         -          -         -          -   12 Juv. Atooth.          -          -      0.2      4.0          -          -         -         -         -          -      5.0          -          -          -     2.0          -         -          -         -          -   13 Avian Predators          -          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -   14 Seabirds          -          -          -          -          -          -         -         -         -          -          -          -   50.0          -         -          -         -          -         -          -   15 Deep demersals          -          -      0.3      3.8          -          -         -    4.8          -          -      7.9          -          -          -  17.0          -     3.1          -         -          -   16 Pollock 1+          -      0.5      5.0   37.4          -   11.9      5.1   20.8   25.2          -      7.3      3.7          -      6.0          -      2.0          -          -         -          -   17 Rockfish          -          -      0.4          -          -      3.0      3.0          -    1.0          -          -          -          -          -     3.0          -         -          -         -          -   18 Baleen Whales          -          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -   19 Salmon Fry 0-12          -          -          -      2.0          -      1.0      3.2          -         -     0.1          -          -          -      2.0          -      2.0          -          -         -          -   20 Nshore Demersal          -      0.5          -      1.3      5.0   23.0   23.4          -    9.0          -      8.0          -     4.0      8.7   11.0          -     0.2          -         -     4.0   21 Squid          -          -      2.0      0.2  30.0   12.0          -    8.0          -          -      2.5      1.0          -      0.7      5.0      4.9   15.8          -         -          -   22 Eulachon          -          -          -      0.2      6.0      5.3      4.8   10.7     4.7     0.1      0.9  14.7     4.0      4.7      1.0      4.3          -          -  30.0          -   23 Sea otters      0.5          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -   24 Deep  Epibent          -          -      9.0   14.1          -          -         -    5.1     4.0          -  35.8  25.0          -          -  15.0  26.0   28.8          -         -   11.0   25 Capelin          -          -      0.5      0.1          -      5.5          -    0.1     5.7      0.1      0.2      4.0          -      1.0      0.5      0.9      0.1          -     1.0          -   26 Adult Herring          -   23.5      3.0          -  25.0   12.0   12.5     2.2     4.0     0.1      0.2      5.1          -          -         -          -         -   20.0          -          -   27 Pollock 0          -          -          -          -          -          -         -    0.3     1.0          -      0.3      1.0          -          -         -      0.2      0.1          -         -          -   28 Shal Lg Epibent          -          -          -      2.4          -          -         -         -         -          -      2.8          -          -          -     1.5  10.3          -          -         -   11.0   29 Invert-eat Bird          -          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -   30 Sandlance          -          -          -      0.1      1.0      1.0          -    0.1     1.5     0.1      0.2      4.3          -  15.0          -      0.6      1.1     8.0      1.0     0.5   31 Juv.  Herring          -          -      2.5   10.3      4.0      9.2   18.6     5.6   40.9     0.1      0.8  26.2          -  36.7      0.5      6.0      3.1   22.0      5.7      0.5   32 Jellies          -          -      1.0          -          -          -         -    5.4          -          -          -          -          -          -         -          -         -          -     1.1          -   33 Deep sm infauna          -          -          -          -          -          -         -    0.4          -          -      2.8          -          -          -  10.5          -     0.3          -         -          -   34 Near Omni-zoo          -          -          -          -          -          -         -         -         -          -          -      0.1     1.0      1.0          -      0.2      0.3     0.3      0.1          -   35 Omni-zooplankto          -          -      7.0          -          -          -     7.7     6.7     3.0          -          -  14.9     1.0      6.2      2.0  35.1   40.0   49.7   16.3          -   36 Shal sm Infauna          -          -          -          -          -          -         -         -         -          -      2.8          -          -          -     0.5          -         -          -         -     1.0   37 Meiofauna          -          -          -          -          -          -         -         -         -          -          -          -          -          -     2.0          -         -          -         -          -   38 Deep Lg Infauna          -          -          -          -          -          -         -         -         -          -          -          -          -          -  10.0          -         -          -         -          -   39 Shal Sm Epibent          -          -      2.0      2.4          -          -         -         -         -          -  10.0          -     4.0      5.3      1.5          -     4.7          -  15.7   70.0   40 Shal lg infauna          -          -          -          -          -          -         -         -         -          -          -          -          -      1.8      5.0          -         -          -         -          -   41 Near Herbi-zoo          -          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -     0.2          -   42 Herbi-Zooplankt          -          -          -          -          -          -     3.8          -         -          -          -          -          -      0.1          -      7.4      2.4          -  28.9     2.0   43 Near Phytoplktn          -          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -   44 Offshore Phyto.          -          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -   45 Macroalgae/grass          -          -          -          -          -          -         -         -         -          -          -          -          -          -         -          -         -          -         -          -   46 Nekton falls          -          -      8.0      2.0          -          -         -    9.1          -          -      2.5          -          -          -         -          -         -          -         -          -   47 InshoreDetritus          -          -          -          -          -          -         -         -         -          -          -          -   28.0  10.8      1.0          -         -          -         -          -   48 Offshr Detritus          -          -          -          -          -          -         -         -         -          -          -          -          -          -  10.0          -         -          -         -          -  Import          -          -      4.5      5.2          -          -         -  20.0          -  99.4  10.0          -          -          -         -          -         -          -         -          -    Appendix 5. (continued). (Predator number corresponds with organism in ‘prey’ column.) 134     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996      134Predator Prey   21    22    23    24    25    26    27    28    29    30    31    32    33    34    35    36    37    38    39    40    41    42        1 Transient Orca       -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -       2 Resident Orca          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -       3 Sharks          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -       4 Halibut          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -       5 Porpoise          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -       6 Pinnipeds          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -       7 Lingcod          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -       8 Sablefish          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -       9 Adult Atooth          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     10 Adult Salmon          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     11 Pac. Cod          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     12 Juv. Atooth.          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     13 Avian Predators          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     14 Seabirds          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     15 Deep demersals          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     16 Pollock 1+          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     17 Rockfish          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     18 Baleen Whales          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     19 Salmon Fry 0-12          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     20 Nshore Demersal          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     21 Squid      0.3          -          -      2.0          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     22 Eulachon      0.1          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     23 Sea otters          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     24 Deep  Epibent      0.4          -      4.0   25.0          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     25 Capelin      0.1          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     26 Adult Herring          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     27 Pollock 0      0.1          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     28 Shal Lg Epibent          -          -   16.0          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     29 Invert-eat Bird          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     30 Sandlance          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     31 Juv.  Herring      0.1          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     32 Jellies          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     33 Deep sm infauna          -          -          -   30.0          -         -         -          -         -          -         -         -     4.0          -         -         -          -         -         -          -          -          -     34 Near Omni-zoo      0.3      0.3          -          -      0.3      0.2     0.2          -         -     0.1      0.1     0.1          -          -         -         -          -         -         -          -          -          -     35 Omni-zooplankto   96.6   49.4          -      5.0   41.3   37.7   27.4          -         -   20.9      9.9  22.9     9.0          -         -         -          -         -         -          -          -          -     36 Shal sm Infauna      0.1          -          -          -          -         -         -   19.0  10.0          -         -         -          -          -         -  15.0          -         -         -          -          -          -     37 Meiofauna          -          -          -      3.0          -         -         -          -         -          -         -         -          -          -         -         -  10.0          -         -          -          -          -     38 Deep Lg Infauna          -          -   10.0      5.0          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     39 Shal Sm Epibent      0.3      0.1          -          -      3.4      8.3          -   79.0  90.0     6.3          -         -          -          -         -         -          -         -     5.0          -          -          -     40 Shal lg infauna          -          -   70.0          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          -     41 Near Herbi-zoo      0.1          -          -          -      0.4      0.3     0.5          -         -     0.5      0.6     0.4          -     0.2     0.2          -          -         -         -          -          -          -     42 Herbi-Zooplankt      1.5      0.2          -      5.0   54.6   53.5   71.9          -         -   72.2   89.4  66.6     9.0   24.8   24.8          -          -  10.0          -          -          -          -     43 Near Phytoplktn          -          -          -          -          -         -         -          -         -          -         -    0.7          -   75.0          -  60.0          -         -  35.0  50.0    100          -     44 Offshore Phyto.          -          -          -          -          -         -         -          -         -          -         -    9.3          -          -  75.0          -          -         -         -          -          -   100     45 Macroalgae/grass          -          -          -      5.0          -         -         -     1.0          -          -         -         -          -          -         -         -          -         -  20.0          -          -          -     46 Nekton falls          -          -          -          -          -         -         -     1.0          -          -         -         -          -          -         -         -          -         -     0.3          -          -          -     47 InshoreDetritus          -          -          -          -          -         -         -          -         -          -         -         -          -          -         -  25.0  45.0          -  39.7  50.0          -          -     48 Offshr Detritus          -          -          -   20.0          -         -         -          -         -          -         -         -   78.0          -         -         -  45.0   90.0          -          -          -          -  Import          -   50.0          -          -          -         -         -          -         -          -         -         -          -          -         -         -          -         -         -          -          -          - ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996                     135Appendix 6. Diagram of the spatial distribution of arrowtooth founder in Prince William Sound (Mark Willette, Alaska Dept. of Fish and Game, personal communication; also see Arrowtooth flounder section). Approximately 56% of the juvenile biomass and 80% of the adult biomass of arrowtooth flounder occurs in southwestern PWS (yellow). The remainder of the juvenile biomass is found in Orca Bay and Port Fidalgo (blue). This diagram is presented here as an example of the types of spatial distribution and habitat information contributed by collaborators for specification of habitat-based Ecospace modelling (see sections on Ecosim and Ecospace). 61.20'         148 W      147 W   146 W                                             61.10'                                                                            61 N                                                                            60.50'                                                                            60.40'                                                                            60.30'                                                                            60.20'                                                                            60.10'                                                                            60 N                                                                            59.50'                   136     ECOPATH MODEL OF PRINCE WILLIAM SOUND, ALASKA, 1994-1996          Commercial Whaling - The Issues Reconsidered FCRR 1993, Vol.1 (1), 36pp; $20  Decision Making by Commercial Fisherman: a Missing Component in Fisheries Management? FCRR 1993, Vol.1 (2), 75pp; $20  Bycatch in Fisheries and their Impact on the Eco-system FCRR 1994, Vol.2 (1), 86pp; $20  Graduate Student Symposium on Fish Population Dynamics and Management FCRR 1995, Vol.3 (2), 33pp; $20  Harvesting Krill: Ecological Impact, Assessment, Products and Markets FCRR 1995, Vol.3 (3), 82pp; $20  Mass-Balance Models of North-eastern Pacific Ecosystems FCRR 1996, Vol.4 (1),131pp; $20  Reinventing Fisheries Management FCRR 1996, Vol.4 (2), 84pp; $20  The Design & Monitoring of Marine Reserves FCRR 1997, Vol.5 (1), 47pp; $20  Assessment of Hong Kong’s Inshore Fishery Re-sources FCRR 1998, Vol.6 (1),148pp; $40  Use of Ecopath with Ecosim to Evaluate Strategies for Sustainable Exploitation of Multi-Species Re-sources FCRR 1998, Vol.6 (2), 49pp; $20  Graduate Student Symposium on Fish Population Dynamics and Management FCRR 1998, Vol.6 (3), 40pp; $20  Trophic Mass-Balance Model of Alaska’s Prince William Sound Ecosystem, for the Post-Spill Pe-riod 1994-1996 FCCR 1998, Vol.6 (4), 143pp; $40  Back to the Future: Reconstructing the Strait of Georgia Ecosystem FCRR 1998, Vol.6 (5), 99pp; $40  Distant Water Fleets: An Ecological, Economic and Social Assessment FCRR 1998, Vol. 6 (6), 111pp; $40  Ecosystem Change and the Decline of  Marine Mammals in the Eastern Bering Sea FCRR 1999, Vol.7 (1), 106pp, $40  Evaluating the Benefits of Recreational Fishing FCRR 1999, Vol.7 (2), 170pp, $40  Back to the Future: Reconstructing the  Hecate Strait Ecosytem FCRR 1999, Vol.7 (3), 65pp, $20   Copies of any of these research reports may be obtained at a cost from $20.00 to $40.00, which includes surface mail. Payment may be made by Credit Card, cheque or money order.  Please contact:  Events Officer Fisheries Centre University of British Columbia   phone:  604 822-0618 2204 Main Mall, Hut B-8    fax:  604 822-8934 Vancouver, V6T 1Z4    E-mail:  events@fisheries.com Canada      Web Site:  http://fisheries.com FISHERIES CENTRE RESEARCH REPORTS 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics