Open Collections

UBC Graduate Research

From the Tropics to the Poles : Ecosystem Models of Hudson Bay, Kaloko-Honokōhau, Hawai'i, and the Antarctic… Wabnitz, Colette; Hoover, Carie 2012

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

Item Metadata

Download

Media
42591-Wabnitz_C_et_al_From_tropics_poles_ecosystem.pdf [ 2.8MB ]
Metadata
JSON: 42591-1.0354242.json
JSON-LD: 42591-1.0354242-ld.json
RDF/XML (Pretty): 42591-1.0354242-rdf.xml
RDF/JSON: 42591-1.0354242-rdf.json
Turtle: 42591-1.0354242-turtle.txt
N-Triples: 42591-1.0354242-rdf-ntriples.txt
Original Record: 42591-1.0354242-source.json
Full Text
42591-1.0354242-fulltext.txt
Citation
42591-1.0354242.ris

Full Text

 ISSN 1198-6727 FROM THE TROPICS TO THE POLES: ECOSYSTEM MODELS OF HUDSON BAY, KALOKO-HONOKŌHAU, HAWAI‘I, AND THE ANTARCTIC PENINSULA Fisheries Centre Research Reports 2012 Volume 20 Number 2     ISSN 1198-6727  Fisheries Centre Research Reports     2012 Volume 20 Number 2  FROM THE TROPICS TO THE POLES: ECOSYSTEM MODELS OF HUDSON BAY, KALOKO-HONOKŌHAU, HAWAI‘I, AND THE ANTARCTIC PENINSULA  Fisheries Centre, University of British Columbia, Canada     FROM THE TROPICS TO THE POLES: ECOSYSTEM MODELS OF HUDSON BAY, KALOKO-HONOKŌHAU, HAWAI‘I, AND THE ANTARCTIC PENINSULA      Edited by Colette C. C. Wabnitz and Carie Hoover            Fisheries Centre Research Reports 20(2) 182 pages © published 2012 by  The Fisheries Centre, AERL University of British Columbia  2202 Main Mall Vancouver, B.C., Canada, V6T 1Z4     ISSN 1198-6727   Fisheries Centre Research Reports 20(2) 2012   FROM THE TROPICS TO THE POLES: ECOSYSTEM MODELS OF HUDSON BAY, KALOKO-HONOKŌHAU, HAWAI‘I, AND THE ANTARCTIC PENINSULA    Edited by Colette C. C. Wabnitz and Carie Hoover   CONTENTS  AN ECOSYSTEM MODEL OF HUDSON BAY, CANADA WITH CHANGES FROM 1970-2009 ........................................... 2 ABSTRACT .............................................................................................................................................. 2 INTRODUCTION ...................................................................................................................................... 2 MATERIALS AND METHODS .................................................................................................................... 3 ECOSYSTEM OVERVIEW ............................................................................................................. 3 MODEL EQUATIONS .................................................................................................................. 4 MODEL PARAMETERS BY FUNCTIONAL GROUP ........................................................................... 5 ECOSIM PARAMETERS ............................................................................................................. 26 FITTING THE MODEL TO DATA ................................................................................................ 30 RESULTS .............................................................................................................................................. 33 BALANCING THE MODEL ......................................................................................................... 33 MONTE CARLO RESULTS ......................................................................................................... 35 FITTING RESULTS ................................................................................................................... 36 DISCUSSION ......................................................................................................................................... 39 CONCLUSIONS ...................................................................................................................................... 40 ACKNOWLEDGEMENTS ......................................................................................................................... 41 REFERENCES ........................................................................................................................................ 42 APPENDIX 1 : MARINE MAMMAL MORTALITY CALCULATIONS ................................................................ 49 APPENDIX 2 : BIRD SPECIES FOUND WITHIN THE HUDSON BAY MODEL AREA ......................................... 50 APPENDIX 3:  FISH FUNCTIONAL GROUPS AND SPECIES INCLUDED IN EACH GROUP ................................. 54 APPENDIX 4: VULNERABILITIES USED TO FIT THE MODEL ...................................................................... 55 APPENDIX 5: MIXED TROPHIC IMPACT RESULTS .................................................................................... 58 APPENDIX 6: COEFFICIENT OF VARIATION (CV) VALUES USED FOR MONTE CARLO ROUTINE .................. 63 APPENDIX 7: MONTE CARLO RESULTS FOR ESTIMATES OF BIOMASS ........................................................ 64 APPENDIX 8: ECOSIM OUTPUT BIOMASS TRENDS ................................................................................... 67 BASELINE TROPHIC RELATIONSHIPS IN KALOKO-HONOKŌHAU, HAWAI‘I ............................................................. 69 ABSTRACT ............................................................................................................................................ 69 INTRODUCTION .................................................................................................................................... 69 MATERIALS AND METHODS ................................................................................................................... 71 STUDY AREA ............................................................................................................................ 71 MODELING APPROACH ............................................................................................................ 72 MODEL PARAMETERS AND FUNCTIONAL GROUPS .................................................................... 73 RESULTS .............................................................................................................................................. 86 DISCUSSION ......................................................................................................................................... 89 DESCRIPTION OF THE KALOKO-HONOKŌHAU SYSTEM .............................................................. 89 TURTLES AT CARRYING CAPACITY ............................................................................................ 91 POTENTIAL THREATS DUE TO URBAN DEVELOPMENT .............................................................. 92 ACKNOWLEDGMENTS ........................................................................................................................... 94 REFERENCES ........................................................................................................................................ 95 APPENDIX 1: SPECIES INCLUDED UNDER EACH REEF FISH FUNCTIONAL GROUP ..................................... 106 THE ANTARCTIC PENINSULA MARINE ECOSYSTEM MODEL AND SIMULATIONS: 1978-PRESENT ......................... 108 ABSTRACT ......................................................................................................................................... 108 INTRODUCTION .................................................................................................................................. 108 METHODS .......................................................................................................................................... 110 MODEL EQUATIONS ............................................................................................................... 110 SPECIES GROUPS .................................................................................................................... 111 ECOSIM INPUT ...................................................................................................................... 137 RESULTS ............................................................................................................................................ 144 MODEL BALANCING ............................................................................................................... 144 FITTING THE MODEL ............................................................................................................. 146 DISCUSSION ....................................................................................................................................... 153 ACKNOWLEDGEMENTS ........................................................................................................................ 155 REFERENCES ...................................................................................................................................... 156 APPENDIX  1 : VULNERABILITIES USED FOR THE FITTED MODEL .......................................................... 166 APPENDIX 2: MIXED TROPHIC IMPACTS .............................................................................................. 169 APPENDIX 3: CV VALUES USED FOR THE MONTE CARLO ANALYSIS ...................................................... 174 APPENDIX 4: MONTE CARLO RESULTS ................................................................................................. 175 APPENDIX 5: MONTE CARLO RESULTS FOR BIOMASS ........................................................................... 177 APPENDIX 6: FITTED MODEL COMPARISONS ...................................................................................... 180 A Research Report from the Fisheries Centre at UBC 182 pages © Fisheries Centre, University of British Columbia, 2012 FISHERIES CENTRE RESEARCH REPORTS ARE ABSTRACTED IN THE FAO AQUATIC SCIENCES AND FISHERIES ABSTRACTS (ASFA) ISSN 1198-6727 From the Tropics to the Poles, Wabnitz and Hoover  DIRECTOR’S FOREWORD This report summarizes the existing knowledge on three ecosystems: Hudson Bay, Canada, Kaloko-Honokōhau, Hawai‘i, and the Antarctic Peninsula, Antarctica. Through the construction of ecosystem models representing these three regions, research from numerous aspects of each ecosystem are pieced together to present a holistic story. While we live in a rapidly changing world, it is important to remember there are many regions where we are still gaining an understanding of basic knowledge.  Research on these ecosystems, from the Arctic to the tropics to the Antarctic, presents different levels of our knowledge. For the Arctic (Hudson Bay) the focus is identifying changes known to be occurring for certain species, and addressing the reasons for those changes in addition to the greater implications to the rest of the ecosystem. In the tropics (Hawai‘i) the construction of a model allows insight into structure and function of the ecosystem focusing on the role of an endangered species, the green sea turtle, and provides a baseline to assess potential future impacts on the ecosystem from coastal development. In the Antarctic (Antarctic Peninsula) ecosystem, environmental changes are explored as they impact a key link in the food web.  While the models presented address localized issues relating to very different regions of the world, the ultimate goal is the same; to increase our understanding of ecosystems as a whole and the different stressors related to each region. With this knowledge, we can formulate better questions for future research, assist in informing managers, and hopefully gain greater insights and understanding of the likely impact of future stressors.  U. RASHID SUMAILA Director and Professor UBC Fisheries Centre December 2012     2  An Ecosystem Model of Hudson Bay, Canada    AN ECOSYSTEM MODEL OF HUDSON BAY, CANADA                                                              WITH CHANGES FROM 1970-20091 Carie Hoover, Tony J Pitcher, and Villy Christensen Fisheries Centre, University of British Columbia, 2202 Main Mall, Vancouver BC V6T 1Z4 Email:c.hoover@fisheries.ubc.ca ABSTRACT An ecosystem model was created for the Hudson Bay region, Canada, for 1970-2010, aiming to identify the threats of global warming to marine mammals. The model presented in detail here includes 40 functional groups and provides estimates for previously unknown parameters such as biomass of fish species, given ecosystem constraints, using the Ecopath with Ecosim modeling framework. The model is tuned to all catch data known for the region. In addition to providing a comprehensive overview of the trophic dynamics within the system, temporal simulations driven by declines in sea ice mimic the changes known to occur to the region. The model captures many dynamics present in the system, while identifying gaps in existing data for future research and as the basis for work simulating climate change and its impacts on the ecosystem.  INTRODUCTION Polar Regions are increasing in temperature faster than temperate areas, with Arctic temperature rising at almost twice the rate of the rest of the world (Arctic Climate Impact Assessment 2004). The fourth International Polar Year (IPY) in 2007-2009 highlighted the need for research to increase our knowledge of the dynamics occurring in Polar areas.  Hudson Bay (HB) is a unique ecosystem. While in location it is considered sub-Arctic, between 50-70˚N, the climate of this system is atypical reflecting high Arctic climate and biogeography, especially when considering higher trophic level animals. For example, polar bears, one of the most famous examples, are found at their lowest latitudinal range in HB, due to the cold winters and the ice available for foraging (Stirling and Parkinson 2006). Moreover, many species present in this ecosystem have adapted to the seasonal ice cycle, from whales occupying the region during the ice free seasons, and seals breeding on the ice, to the ability of smaller zooplankton to survive winter months using nutrients frozen within the sea ice (Poltermann 2001; Stewart and Lockhart 2005). While surface temperatures in Hudson Bay have increased 0.5-1.5˚C during 1955-2005 2005 (Hansen et al. 2006), sea ice decreased by 2000±900 km2 year-1 between 1978 and 1996 (Parkinson et al. 1999). These changes combined with a longer ice free season (Gagnon and Gough 2005) have unknown consequences for the marine ecosystem. There is a greater need than ever to utilize the available data to discover new information within HB and other polar regions. HB, being relatively unstudied compared to other parts of the Arctic and Antarctic, became the focus for an ecosystem model as part of the Global Warming and Arctic Marine Mammals (GWAMM) project as funded by IPY (International Polar Year) and DFO (Department of Fisheries and Oceans). The aim of the ecosystem model presented in this report is to gain a deeper understanding of the trophic links within the Hudson Bay, particularly those impinging on marine mammals, and to understand how the ecosystem structure might change due to future climate change.                                                   1 Cite as: Hoover, C., Pitcher, T.J., and Christensen, V. (2012) An Ecosystem Model of Hudson Bay, Canada and Changes from 1970-2009, p. 2-68. In: Wabnitz, C.C.C. and Hoover, C. (eds.) From the tropics to the poles: Ecosystem models of Hudson Bay, Kaloko-Honokōhau, Hawai‛i, and the Antarctic Peninsula. Fisheries Centre Research Reports 20(2). Fisheries Centre, University of British Columbia [ISSN 1198-6727]. From the Tropics to the Poles, Wabnitz and Hoover  MATERIALS AND METHODS Ecosystem Overview The Hudson Bay ecosystem has been an integral part of Canadian history, most notably in the 18th century with the expansion of the fur trade by Hudson’s Bay Company (Stewart and Lockhart 2005). Prior to this Thule and Dorset cultures had survived for thousands of years by hunting marine mammals. Presently, Inuit and Cree populations inhabit the coast of Hudson Bay, along with additional communities further north in the Canadian archipelago. The majority of Inuit and Cree reside in Nunavut and the Nunavik portion of Quebec (which represents the top third of the province, and is comprised of First Nations).  The greater HB complex often includes Hudson Bay, James Bay (JB), Foxe Basin (FB) and Hudson Strait (HS) (figure 1). This system is one of the largest bodies of water in the world to freeze over every winter and open up every summer. HB and JB are both categorized by shallow, less productive waters, with large inputs of freshwater from rivers in the spring. Conversely, Foxe Basin and Hudson Strait have more mixing with the Labrador Sea (Straneo and Saucier 2008), and are thought to be an important sea ice chokepoint for Hudson Bay, ultimately determining which marine species have access to the region (Higdon and Ferguson 2009).  There is a large gradient regarding the types of lower trophic level species found from the north to southern HB, although marine mammals are able to utilize the whole region. Selection of the model area was based on use patterns of marine mammals as their data is more prevalent compared to fish and plankton species, and because modelling aims to focus on the impact of climate change on marine mammal species. James Bay was included in the model area due to its similarity to southern HB. Belugas from two of the three stocks found within HB reside in or near JB in the summer months; therefore this region was believed to be important. Hudson Strait and Foxe Basin were excluded from the model area, as the model area was already quite large and incorporation of these areas was not believed to be essential to Figure 1: Greater Hudson Bay region and associated communities. 4  An Ecosystem Model of Hudson Bay, Canada    understanding the dynamics within HB. For the remainder of this paper, referral to Hudson Bay will include James Bay.   The time period for the model was selected to be 1970. This was chosen as few studies were completed prior to the 1960s regarding estimates of marine species. In addition decreases in sea ice and rising temperatures have been recorded over the last 50 years (Gagnon and Gough 2005), thus making for an interesting time to examine the ecosystem. Finally, there was a lack of data pre-1970, indicating no additional information regarding the ecosystem would be gained by expanding the modelling temporally.  Model Equations  The basic Ecopath model requires various parameters as input; biomass, production, consumption, and ecotrophic efficiency (Christensen et al. 2005). Production and consumption values are entered as a ratio to the biomass of a species group (i.e. production/consumption). The master equations (eq. 1-3) to this modelling method are: (1) Production=predation + fishery+ biomass change+ net migration+ other mortality (2) Consumption= production + unassimilated food + respiration (3) EE= 1- other mortality  Biomasses for all groups, where available, were based on the abundance of individual species or species groups multiplied by an average weight per individual and divided among the model area. For Hudson Bay and James Bay the estimated area was nearly 900,000 km2 (Legendre et al. 1996). For each functional group (or species group) one parameter may be left missing to be solved by the program. Through the use of linear equations, and trophic links, these missing parameters may be solved for. Trophic links are incorporated in the model through the use of the diet matrix.  Temporal simulations are then generated in Ecosim using equation 4;  																			ሺ4ሻ																												݀ܤ௜ ݀ݐ⁄ ൌ ݃௜෍ܳ௝௜ െ௝෍ܳ௜௝ ൅ ܫ௜ െ ሺܯ ௜ܱ ൅ ܨ௜൅݁௜ሻܤ௜௝ where	݀ܤ௜ ݀ݐ⁄  represents the change in biomass (B) for group i over the time interval t, with starting biomass	ܤ௜. 	 ௜݃ 	represents the net growth efficiency (production/consumption ratio ,the ∑ ܳ௝௜௝  is the total consumption on group i, and ∑ ܳ௜௝௝  is the predation of all predators on group i. ܯ ௜ܱ represents the other mortality term (for mortality associated with old age), ܨ௜ is the fishing mortality rate, ܫ௜ is the immigration rate, ݁௜	is the emigration rate, with the combined term ሺܤ௜ ∙ ݁௜ െ ܫ௜ሻ as the net migration rate.  The consumption rate of a group, Qij, is based on the foraging arena theory where the biomass Bi is further divided into vulnerable and invulnerable proportions to group i’s predators (Walters et al. 1997), and the transfer rate between these two states. Vulnerabilities are set within the model with values =2 indicating a mixed interaction between predator and prey, >2 resulting in a bottom up interaction, and <2 resulting in a top down interaction. Vulnerabilities are estimated through an automated search routine (Fit to time series), and user determined values. Monte Carlo simulations were run on the fitted model, using a pedigree ranking, whereby input parameters are assigned a coefficient of variation (CV) based on the quality of input data from the pedigree ranking from Ecopath with Ecosim version 5 (Christensen et al. 2005). 1000 simulations were run to estimate ranges of biomass and P/B values of all functional groups.   From the Tropics to the Poles, Wabnitz and Hoover  Model Parameters by Functional Group Marine Mammals All marine mammals which inhabit the model area were included in the model. In addition, many species have been shown to be representatives from genetically distinct stocks, and therefore have been split into individual functional groups. For example, there are three stocks of polar bears within the model area, each with differing population trends and hunting quotas, and were therefore considered different stocks and functional groups within the model. Four species of cetaceans (bowhead whales, narwhals, belugas, and killer whales) are seasonal residents in Hudson Bay. For these functional groups their impact on the ecosystem is relative to the amount of time spent in the area and the proportion of annual feeding occurring during their time in Hudson Bay. A weighted biomass was designated to each of these groups to represent their respective impact on the ecosystem, so that if a group of whales resided in Hudson Bay half of the year and half of their feeding occurred during this time, then their weighted biomass would be half of the total population biomass (50%) to account for this. Individual estimates are given within functional group parameters. For all marine mammal groups the following equations were used to calculate input parameters (parameters for all marine mammals are listed in table 1). Biomass was calculated by multiplying the number of individuals by average weight of individuals (in tonnes), then divided by the model area (km2). Mortality rates (P/B ratios) were calculated for each species using the life table based on natural mortality (Barlow and Boveng 1991), and compared to published values where available (full equations for P/B calculations are available in appendix 1.). Mortality from hunting was calculated as the biomass harvested/total biomass, and was added to the natural mortality to give the final P/B ratio. Q/B: Consumption (Q/B) was calculated using equation 5 (Hunt et al. 2000; Guénette 2005);  ሺ5ሻ																																												ܧ ൌ ܽܯ଴.଻ହ  where E in the energy required per day (Kcal/day), M is the mean body weight (in Kg) and a is a coefficient representing each group of marine mammals (a=320 for otariids, 200 for phocids, 192 for mysticetes, 317 for odontocetes, and 320 for sea otters). Energy contents of food items was provided by various authors as summarized in Cauffopé and Heymans (2005).            6  An Ecosystem Model of Hudson Bay, Canada    Table 1: Input parameters for marine mammal functional groups Species Pop Size Source Mean Weight (Kg) Source Longevity (Years) Source Mortality (calculated) Hunting Mortality Model P/B Calculated Q/B Model Q/B Polar Bear WHB 1200 (Lunn et al. 2002) 300 (Stirling and Parkinson 2006) 25 (Stirling 2002) 0.096 0.033 0.129 3.029 2.080 Polar Bear SHB 1000 (Lunn et al. 2002) 300 (Stirling and Parkinson 2006) 25 (Stirling 2002) 0.096 0.058 0.154 3.109 2.080 Polar Bear Foxe 3000 (Aars et al. 2005) 300 (Stirling and Parkinson 2006) 25 (Stirling 2002) 0.096 0.024 0.120 2.849 2.080 Killer Whale† 20 (Higdon and Ferguson 2009) 4689 (Ford 2002) 80 (Ford 2002) 0.048 0.051 0.151 4.998 4.998 Narwhal† 2710 (Richard 1991) 1300 Heide-Jørgensen 2002) 115 (Garde et al. 2007) 0.083 0.008 0.084 18.696 26.182 Bowhead† 64 Higdon 2009 unpublished data 31076 (Trites and Pauly 1998) 200 (George et al. 1999) 0.018 0.003 0.021 5.475 5.475 Walrus N 2500 (Mansfield and St Aubin 1991) 1976 estimate 1037.5 (Kastelein 2002) 35 (Kastelein 2002) 0.141 0.031 0.172 41.238 47.123 Walrus S 500  (Richard and Campbell 1988 : COSEWIC 2006) 1037.5 (Kastelein 2002) 35 (Kastelein 2002) 0.088 0.009 0.097 29.560 33.778 Bearded Seal 15000 (Lunn et al. 1997) 275 (Kovacs 2002) 25 (Kovacs 2002) 0.131 0.045 0.176 13.848 14.262 Harbour Seal 1000 Assumed 76 (Burns 2002) 29.5 (Trites and Pauly 1998) 0.123 0.002 0.125 18.612 18.612 Ringed Seal 600,000 (Smith 1975) 42.5 (Trites and Pauly 1998) 43 (Miyazaki 2002) 0.15 0.008 0.158 16.050 17.272 Harp seal 8000 Assumed (Ferguson pers. Comm) 130 (Lavigne 2002) 30 (Lavigne 2002) 0.112 0.014 0.126 15.660 15.660 Belgua E† 4200 (Hammill 2001 ; Gosselin 2005 ; Hammill et al. 2009a) 725 (DFO 2002b; NAMMCO 2005b) 50 (Harwood et al. 2002; Stewart et al. 2006)  0.044 0.032 0.066* 21.448 21.448 Belgua W† 50,000 (COSEWIC 2004a; NAMMCO 2005b) 725 (DFO 2002b; NAMMCO 2005b) 50 (Harwood et al. 2002; Stewart et al. 2006) 0.0587 0.005 0.064 16.713 16.713 Beluga James† 1842 (Gosselin et al. 2002) 725 (DFO 2002b; NAMMCO 2005b) 50 (Harwood et al. 2002; Stewart et al. 2006) 0.057 0.019 0.087* 16.623 16.623 † Narwhal, Killer whale, bowhead, and all beluga biomasses were adjusted to 50% to account for roughly 50% of their time spent in the model area *  The P/B for Eastern Belugas and James Belugas also account for migrations which were added in the fitting process        From the Tropics to the Poles, Wabnitz and Hoover  Polar Bears (Ursus maritimus) Three of the nineteen polar bear populations (Paetkau et al. 1999) overlap with the Hudson Bay ecosystem model area; the Western Hudson Bay population, the Southern Hudson Bay population, and part of the Foxe Basin population (figure 2). These three populations were included in the model under different functional groups corresponding to each population (Western Hudson Bay, Southern Hudson Bay, and Foxe Basin). Being at the southern range of their limits in HB, climate change is believed to be an important factor in determining the health of these populations. Since polar bears rely on ice for foraging, extension to the ice free summer caused by melting is believed to increase nutritional stress. In addition, because these southerly populations already experience longer summers than their northern counterparts, they are thought to be more vulnerable to declines in sea ice (Stirling and Derocher 1993; Stirling et al. 1999). The Foxe Basin (FB) and Western Hudson Bay (WHB) populations are believed to be declining, while there have not been enough surveys to determine trends in the Southern Hudson Bay (SHB) stock (Aars et al. 2005). In addition each population is subjected to different hunting pressures depending on the communities within their respective ranges. While diets vary among populations, ringed seals are the most important food item in all polar bear populations, followed by bearded and harp seals (Peacock et al. 2010). Polar bears have also been known to take walrus, beluga, narwhal, seabirds, and waterfowl (Stirling 2002). Scat analysis of western and southern HB polar bears form the late 1960s indicated foraging on birds (primarily from the family Anatidae- ducks, swans, and geese), mussels, urchins, other unidentifiable invertebrates, and berries in the late summer and autumn (Russell 1975; Derocher et al. 1993).  Although it is likely that these prey items are also consumed by the Foxe Basin population, it is believed the WHB and SHB may consume a greater portion of birds, invertebrates, and plants in their diets.  Polar bears were traditionally hunted for food and clothing, a tradition which still exists today. Quotas have been imposed on each of the stocks by corresponding jurisdictions in Nunavut, Ontario, Manitoba, and Quebec (Peacock et al. 2010). 1. Western Hudson Bay Polar Bears The western HB polar bear population has been declining since 1981. The decline is believed to be caused by a lengthening of the ice free season (summer), which has led to increased nutritional stress (Stirling et al. 1999). The increased open water season is correlated with poor condition especially in female polar Figure 2: Delineation of the Western Hudson Bay (WHB), Southern Hudson Bay (SHB), and Foxe Basin (FB) populations of polar bears within Hudson Bay. Reprinted from Stirling et al. (1999).  8  An Ecosystem Model of Hudson Bay, Canada    bears (Stirling et al. 1999). The population was estimated at 1200 bears based on an estimate from 1987 (Lunn et al. 2002), giving a biomass for the region of 0.00046 t·km-2. In 2004 the population is believed to have dropped to 935 animals (Aars et al. 2005). An average catch of 44 bears during the 1980s (Lee and Taylor 1994) has since increased slightly to 46.8 bears for the 1999-2004 period (Aars et al. 2005). The 2005 quota for the WHB polar bear population was 56 bears (Aars et al. 2005). Diet was set to 1% polar bears (Western Hudson Bay), 0.5% northern walrus, 12.5% bearded seals, 0.1% harbour seals, 61.9% ringed seals, 3% harp seals, 10% western beluga whales, 2% seabirds, 1% each echinoderms and bivalves, 7% other benthos.  2. Southern Hudson Bay Polar Bears  The SHB polar bear population was estimated at 1000 bears in the 1980s (Lunn et al. 2002), giving a biomass of 0.000383 t·km-2 for the entire region. As there have been no estimates of this population since, therefore the estimate of 1000 bears was used for the starting 1970s biomass. The diet for SHB polar bears was set to: 1% SHB polar Bears (to account for cannibalism), 0.5% southern walrus, 12.5% bearded seals, 0.1% harbour seals, 62.4% ringed seals, 3% harp seals, 0.5% eastern belugas, 6.5% James Bay belugas, 7% seabirds, 2% echinoderms, 2% bivalves, and 2.5% other benthos. The average catch of SHB polar bears for the 1980s was 68 (Lee and Taylor 1994), and with no previous records available, this values was assumed to be the catch for 1970.  3. Foxe Basin Polar Bears  The FB polar bear population has shown a decrease from 3000 bears (1970s) to 2100 (1996), and then a slight increase to 2300 in 2004 (Aars et al. 2005). This population is not fully within the model limits so the 1970s abundance would yield a biomass of 0.000986 t·km-2, however, it was assumed only 20% of the population was geographically located within the model area, so the biomass was adjusted to 0.000197 t·km-2.  Average catches for the 1980s were 142 bears (Lee and Taylor 1994). This value was used as the catch in 1970, although again it was also adjusted to 20% of its value to account for the percentage taken from within the model area. The diet for FB polar bears is believed to contain less birds and invertebrates and more seals (Russell 1975) and was therefore set to 0.5% FB polar bears, 20% bearded seals, 1% harbour seals, 59.5% ringed seals, 4% harp seals, 8% western Belugas, 2% seabirds, 1.5% echinoderms, 1.5% bivalves, and 2% other benthos.  Killer Whales (Orcinus orca)  There has been an observed increase in the number of killer whales present in Hudson Bay since the 1950s, which has been linked to the decreasing ice cover in the region (Higdon and Ferguson 2009). Killer whales move into Hudson Bay through Hudson Strait in the summer when the ice has melted enough to allow them to travel through, and they leave before the annual freeze-up. It is believed they travel into the area following other marine mammal species as food, although a determined ecotype has not been established for these animals. Inuit knowledge suggests killer whales were not present prior to the mid-1900s but are now observed on a regular basis (Gonzalez 2001). A photo identification project established in 2005 has identified 67 unique individuals in the Eastern Arctic (Peterson et al. 2009).  The 1970s population was set to 20 individuals or a biomass of 0.000025 t·km-2 based on the conservative population estimate for the 2000s of at least 67 individuals and sightings, which have increased nearly fivefold since the 1970s (Higdon and Ferguson 2009). Although killer whales only enter HB during the ice-free season, it was assumed that, for the proportion of the population that do, they feed completely on the species in the model area. Therefore no adjustments to the biomass were made. Reported observations of predation consist of marine mammals, although not enough research has been completed to identify this population of killer whales as marine mammal consumers. In addition, reports from killer whales in other areas of Canada have stated observations of whales eating fish (Lawson et al. 2007; Higdon and Ferguson 2009). The diet was therefore set primarily to marine mammals with some From the Tropics to the Poles, Wabnitz and Hoover  fish and birds being consumed; 8% narwhal, 2.5% bowhead, 6% walrus (3% each north and south walrus), 13% bearded seal, 1.5% harbour seal, 33% ringed seal, 3% harp seal, 22% beluga (1%eastern, 16% western, 5% James Bay), 3% seabirds, 0.5% Atlantic Salmon, 3% gadiformes, 2% sculpins/zoarcids, 0.5% sharks/rays, 1% other marine fish, and 1% cephalopods (Gonzalez 2001; Higdon 2007; Higdon and Ferguson 2009). Based on increased sightings in Higdon and Ferguson (2009) for Hudson and James Bays, sightings of killer whales was assumed to be directly proportional to the number of killer whales present. A review of literature by Higdon (2007) summarized reported kills of killer whales from 1957 onwards in the eastern Canadian Arctic. Since killer whales are occasionally harvested, hunting mortality for the first year was set intentionally low; to the equivalent of half the biomass of one whale to give a hunting mortality of 0.103 year-1. This combined with the natural mortality led to a P/B of 0.151 year-1 to be used in the model.  Narwhal (Monodon monoceros)  The Northern Hudson Bay stock of narwhal is the smallest of three narwhal stocks (Northern Hudson Bay, Baffin Bay, and Greenland Sea) in the Arctic (COSEWIC 2004b). Narwhals are found near the Repulse Bay area of Hudson Bay in the summer months, and migrate to the Labrador Sea for the winter, spending roughly half of the year within the HB model area. Although the wintering area for the Hudson Bay stock and the Baffin Bay stock overlap, summer site fidelity indicates they are different stocks (Westdal et al. 2010).  The stock for Hudson Bay was estimated to be 1355 individuals in 1984 (Richard 1991), however this analysis did not account for submerged animals during the sampling, and should be doubled (to 2710) to more accurately represent the population. An estimate of 1780 whales for the population in 2000, also under-representative due to diving animals was corrected to 3500 whales, which is believed to be a more accurate value (COSEWIC 2004b). Biomass and catches were adjusted to 50% of original values to accommodate for time spent and feeding outside of the model area.  Narwhal diets in HB are thought to be focused on Arctic cod, squid, and crustaceans, also including demersal species and invertebrates (Heide-Jørgensen 2002; COSEWIC 2004b; Stewart and Lockhart 2005). The diet was set to 1% Arctic char, 1% Atlantic salmon, 25% gadiformes, 15% sculpins/zoarcids, 12% capelin, 10% other marine fish, 2% brackish fish, 10% cephalopods, 5% macro-zooplankton, 4% euphausiids, and 15% crustaceans.  Bowhead (Balaena mysticetus)  The eastern Canadian Arctic bowhead whales are one of two populations worldwide, with the other being in west Greenland. Previously the Canadian population was believed to be two stocks (George et al. 1999), although genetic sampling has shown not to support this idea suggesting whales are from the same stock (Ferguson 2007). Bowheads are the largest marine mammals within the HB ecosystem, with weight estimates ranging from 54000kg up to 68,000 kg or higher for adult individuals (Rugh and Shelden 2002; American Cetacean Society 2004) and can live for over 200 years (George et al. 1999). They have been an important source of food for historic cultures located in Hudson Bay starting with the Thule near 1000 AD (Higdon 2008). Annual migrations coincide with the ice-free season in HB, where whales move into HB around April to May and leave in September. Although the population has been estimated to be as high as 625 individuals in the 1860s for the HB region, it had dropped as low as or lower than 100 individuals in the late 1800s to early 1900s due to commercial whaling. Since reaching a low in the early 1900s the population has increased with model estimates of 300-400 whales (Higdon 2008 unpublished data). Survey data put the recent HB portion of whales at a minimum of 75 (not accounting for submerged animals at the time of the study) while the Foxe Basin portion of the study identified to be between 256-284 (again not accounting for submerged animals) whales in 1994 (Cosens and Innes 2000). These are now believed to be from the same stock with differing summering grounds, and some sex segregation with mostly cow calf pairs in HB (Higdon and Ferguson 2010). 10  An Ecosystem Model of Hudson Bay, Canada    The historical model estimates the 1970s population to be 319 whales, and it was assumed that roughly 20% of this population will enter Hudson Bay, as based on a 1994 survey where there were 75 whales in HB and 284 in Hudson Strait observed (DFO 1999), giving an estimate of 64 whales. The biomass was then set to 0.0109 t·km-2.  The diet of bowhead whales is believed to consist primarily of copepods and euphausiids with other zooplankton (mysids, gammerid amphipods) and benthic crustaceans being consumed (Lowry et al. 1987; Rugh and Shelden 2002). The diet was set to 5% macro-zooplankton, 30%  euphausiids, 45% copepods, 5% crustaceans, 1% other meso-zooplankton, 5% micro-zooplankton, 2% marine worms, 1% echinoderms, 1% bivalves, and 5% other benthos.  Atlantic Walrus (Odobenus rosmarus) Walrus are year round inhabitants of HB, surviving the winter on the ice. They utilize the sea ice as a platform for breeding, and rely on polynyas in order to feed throughout the winter (Stirling 1997; NAMMCO 2005a). Two of the five recognized stocks (figure 3) of walrus are located partially or fully within HB; the south and east Hudson Bay stock which is completely contained in the model (referred to as Walrus South in the model), and the Hudson Bay-Davis Strait stock (referred to as Walrus North in the model) where the lower portion of the range reaches into the northern part of the model area (DFO 2002a; COSEWIC 2006). There are no complete stock assessments for any of the four walrus stocks, however estimates are presented for each of the HB stocks (DFO 2002a). These stocks were split into two functional groups as they are hunted by different communities, and have different dietary habits.  Figure 3: Ranges of Atlantic Walrus stocks reprinted from (Stewart 2008). Modeled stocks include the walrus N functional group (stock #3 on map) the Northern Hudson Bay-Davis Strait stock and the walrus S (stock #2 on map) known as the Southern and Eastern Hudson Bay stock. No other walrus stocks are used in the model. From the Tropics to the Poles, Wabnitz and Hoover  1. Walrus N The Walrus North species group represents the Hudson Bay- Davis Strait stock. This population has been estimated to contain 3000-4000 animals in the mid 1970s (Richard and Campbell 1988). This estimate represented the population within the entire stock range. However a 1976 survey for the Southampton/ Coates Islands region of northern Hudson Bay estimated 2370 animals in this smaller area (Mansfield and St Aubin 1991). The population within the model was set to 2500 animals, a conservative estimate, or 0.00274 t·km-2 to represent the animals found within the model area from this population.  The diet for these animals consists mainly of benthic invertebrates (bivalves, gastropods, holothurians, polychaetes, and brachiopods), with bivalves contributing to nearly half the diet by weight (Fisher and Stewart 1997; Kastelein 2002; Born et al. 2003). Within the model, the diet was set to: 2% gadiformes, 1% sculpins/zoarcids, 3% other marine fish, 6% crustaceans, 10% marine worms, 25% echinoderms, 40% bivalves, 13% other benthos.  2. Walrus S The Walrus South functional group represents the south and eastern HB stock, which is completely contained within the model area. The population has been estimated to be roughly 410 animals in the late 1970s from surveys at 2 locations in southern HB (310 and 100 walruses), although the reliability of this estimate has been questioned (Richard and Campbell 1988; COSEWIC 2006). Due to lack of better estimates a value of 500 animals was used for the 1970s starting biomass. Although there are no complete surveys, hunters have reported fewer walruses being observed than in the past (DFO 2002a), indicating a declining population. The biomass was set to 0.001 t·km-2.  This stock has been shown to be feeding at higher trophic levels than the other walrus group through stable isotope analysis. While these walruses do still consume bivalves and other invertebrates, they are also feeding on ringed seals and occasionally bearded seals (Muir et al. 1995; Muir et al. 2000). The diet was set to 0.1% bearded seals, 3.9% ringed seals, 8% gadiformes, 1% sculpins/ zoarcids, 5% other marine fish, 7% crustaceans, 10% marine worms, 15% echinoderms, 40% bivalves, 10% other benthos.  Bearded Seal (Erignathus barbatus) Bearded seals are year round inhabitants, using the pack ice and sea ice to haul out. They tend to be found near polynyas or other areas with open access to the water during the winter, and generally inhabit areas with a depth of 200m or less for foraging (Angliss and Outlaw 2006). There have been no studies to suggest there is more than one stock of bearded seals in HB, and although there is no estimate for all bearded seals in HB, surveys have been conducted for the western portion of HB. Lunn et al. (1997) estimated 12900 and 1980 bearded seals for the western portion of HB in 1994 and 1995 respectively based on aerial surveys. It is believed the conditions of the survey played a large role in the discrepancies between estimates. The population for the 1970s was set to 15000 bearded seals for the entire model area or 0.0037 t·km-2, slightly higher than the 1995 estimate. This was set as a conservative estimate for the entire region as there are no known trends for bearded seals, and the surveys did not cover the entire region. It is believed that there may be declines in the bearded seal population as they are a prey item for polar bears, and declining polar bears (Western HB and Foxe Basin) have been shown to be declining possibly because of decreased ringed and bearded seals (Lunn et al. 1997). Hunting of bearded seals is not regulated, with few studies on estimates of numbers hunted (see fisheries section). Bearded seals are benthic feeders with bivalves and crustaceans being the most abundant items in the diet, but fish contributing the highest percent of weight (Smith 1981; Finley and Evans 1983).  Shrimp are more important to newly weaned seals, while adult diets are most likely focused on clams (Young et al. 2010). The diet was set to 3% Arctic char, 2% Atlantic salmon, 20% gadiformes, 5% sculpins/zoarcids, 17% capelin, 4% sandlance, 5% other marine fish, 2% brackish fish, 1% cephalopods, 1% macro-zooplankton, 25% crustaceans, 2% marine worms, 8% echinoderms, 5% other benthos.   12  An Ecosystem Model of Hudson Bay, Canada    Harbor Seal (Phoca vitulina) Harbour seals in Hudson Bay are known to reside in the marine environment as well as lakes which drain into HB (Mansfield 1967a; Smith et al. 1996). The lake seals are not thought to migrate into the marine environment, and are therefore excluded from the model. Although there are no estimates for harbour seals in Hudson Bay, freshwater populations have been estimated between 100-600 seals for specific regions such as Lacs des Loups Marins, Quebec (Smith and Lavigne 1994). Harbour seals are thought to be one of the least abundant seals in HB therefore the abundance was set to 1000 seals or 0.001 t·km-2 (Ferguson pers. comm.).  The diet of harbour seals consists primarily of benthic fish, invertebrates, squid, and crustaceans (Bigg 1981). For the model the diet was set to 10% gadiformes, 8% sculpins/zoarcids, 20% capelin, 20% sandlance, 10% other marine fish, 6% brackish fish, 2% cephalopods, 2% macro-zooplankton, 2%  euphausiids, 10% crustaceans, 3% marine worms, 3% echinoderms, and 4% other benthos.  Ringed Seal (Pusa hispida) Ringed seals are the most abundant seals with a year round distribution in HB. Tagging studies show their ability to travel around Hudson Bay in a matter of weeks. However, seals tagged within Hudson Bay have not been shown to leave the region during the duration of the tagging study (Luque and Ferguson 2008). Because these seals have been shown to travel large distances around HB, all ringed seals in the model area were considered one stock. Recent studies estimated the population size at 73170 in 2007 and 33701 in 2008 for the western portion of HB (DFO 2009) representing only a small portion of the model area. Densities estimated varied from 0.97±0.06 seals km-2 in 2007 to 0.49±0.04 seals km-2 in 2008 for western HB ranging from Arviat to Churchill (Chambellant and Ferguson 2009). If seals were distributed evenly throughout the area the population estimate would range between 450000 and 900000 seals. 1975 estimates from projected population at 61000 seals for James Bay and 455000 from Hudson Bay (Smith 1975). The population for the 1970s was set to 600000 seals, or 0.0469 t·km-2. In general ringed seals feed primarily on Arctic cod and other pelagic fish along with amphipods (DFO 2009). In the Baffin Bay region the diet is dominated by Arctic cod and Polar cod (Holst et al. 2001), but in HB sandlance, euphausiids, and capelin are the most frequent (Chambellant 2010). The diet for Hudson Bay was set to: 18% gadiformes, 10% sculpins/zoarcids, 20% capelin, 30% sandlance, 8% other marine fish, 2% cephalopods, 2% macro-zooplankton, 2% euphausiids, and 8% crustaceans.  Harp seals (Phoca groenlandica) Harp seals are the least abundant of the seal species found in Hudson Bay, although there are no estimates for the abundance in this region. They enter Hudson Bay through Hudson Strait after the break-up of ice in the summer from the Gulf of St Lawrence and south eastern Labrador and leave the area before the freeze up in the fall (Stewart and Lockhart 2005). Population estimates for harp seals in Newfoundland in the 1970s were between 700000 to 1.5 million (Lavigne 1979), however in addition to summering in HB, many animals move to Lancaster Sound, Baffin Bay, Hudson Strait, or Foxe Basin (Mansfield 1967b). For the model the population within HB was estimated to be 8000 (Ferguson pers. comm.) or 0.001 t·km-2. The diet of harp seals from Hudson Strait consists primarily of capelin, and is likely to be similar to the diet of seals within Hudson Bay. Other fish and invertebrate species found from stomach contents were: Arctic cod, sculpin, flatfish, rock cod, mysids, crustaceans, decapods, and other invertebrates (Beck et al. 1993). The diet was set to: 2% Atlantic salmon, 2% gadiformes, 1% sculpins/zoarcids, 86% capelin, 5% other marine fish, and 4% crustaceans.  Beluga (Delphinapterus leucas) Stocks of beluga whales are not fully known for the Hudson Bay region. The North Atlantic Marine Mammal Commission suggests there are 6 groups of belugas within Hudson Bay (NAMMCO 2005b), while genetic studies suggest there are most likely two or three (de March and Postma 2003), based on From the Tropics to the Poles, Wabnitz and Hoover  where whales are hunted or spend a majority of their time. De March and Postma (2003) demonstrate that some belugas harvested from Sanikiluaq are genetically different from the eastern HB and western HB populations. In addition it is possible that belugas harvested from Churchill are also a different stock, although this was not confirmed through genetics. Tagging studies have identified mixing between these populations, making divisions more difficult (Richard and Orr unpublished manuscript as cited in Stewart and Lockhart 2005). For the model 3 functional groups of belugas were created to represent all populations within Hudson Bay: Eastern HB Beluga, Western HB Beluga, and James Bay Beluga. Although mixing between these groups is not well known, for modelling purposes it was assumed there are three separate stocks. As belugas do not spend the winter in HB, the biomass and catches were adjusted to 50% to account for 6 months within the model area.  The general diet of belugas has been noted as consisting primarily of fish species (with pelagic fish being important), benthic invertebrates and squids (Pauly et al. 1998). In the Beaufort sea belugas feed primarily of cod (Loseto et al. 2009), while west Greenland belugas consume squid, molluscs, and myctophids in addition to cod (Heide-Jørgensen and Teilmann 1994). Other noted prey items include crustaceans, worms, and sculpins (Stewart and Lockhart 2005), with capelin as an important component to the diet of eastern and James Bay belugas (Kelley et al. 2010). 1. Beluga East HB Belugas residing in eastern Hudson Bay are considered part of the Ungava and Hudson Bay stock, which is currently listed as endangered by COSEWIC (NAMMCO 2005b). The eastern HB population winters in northern Labrador and makes it migration past Ungava Bay and down the eastern coast of HB to its summer location ranging from Kuujjuaraapik to Inukjuak (DFO 2001). There appears to be a strong genetic basis for designating belugas of Eastern Hudson Bay as a separate population and increasingly good evidence that they contribute to the harvests in Nunavik communities as far as Ungava Bay (COSEWIC 2004a). Areal transect surveys have shown varying trends in the population (Gosselin 2005; Gosselin et al. 2009), however the general trend from surveys and modelling is the population has declined from roughly 4000 whales in 1985 to 2000-3100 whales in 2008 (Hammill 2001; Gosselin 2005; NAMMCO 2005b; Hammill et al. 2009). These declines are thought to be caused primarily by hunting, although noise pollution, river dams, and environmental pollution are also considered factors (DFO 2008). This population was listed as threatened by COSEWIC in 1988, and elevated to endangered status in May 2004 (COSEWIC 2004a). Inuit communities have noted many of the rivers previously utilized by belugas along Hudson Strait and eastern Hudson Bay are no longer used. They believe noise is keeping the whales further offshore in these areas (COSEWIC 2004a). The biomass for the 1970s population was set to 0.00207 t·km-2 or 2100 whales (4200 whales at 50% of the time in the model area). The diet was set to: 2% Atlantic salmon, 8% gadiformes, 10% sculpins, 10% capelin, 5% cephalopods, 2% brackish fish, 15%  euphausiids, 8% copepods, 17% crustaceans, 8% marine worms, and 15% other benthos.  2. Beluga West HB The western Hudson Bay beluga population arrive through Hudson Strait to Churchill, Nelson, and the Seal river estuaries through the summer (COSEWIC 2004a). This population appears to be relatively abundant, although surveys have been sporadic (i.e. 1987 and 2005). COSEWIC (2004a) has designated this population as special concern due to potential substantial removals by hunting throughout its range and concerns with hydroelectric dams and shipping. Estimates show the population as stable. Earlier surveys in 1985 and 1987 estimated the population at 23000 and 25100 whales respectively, while not accounting for submerged animals at the time of the survey (COSEWIC 2004b; NAMMCO 2004b). A 2004 estimate of 57300 whales suggests the population has not changed, as the uncorrected number from this survey is similar to the uncorrected abundances from previous studies (Richard 2005). The 2004 survey also identified an additional 1300 animals along the Ontario coast and 700 along northern HB, however it was not known what stock these whales belonged to. Little genetic testing has occurred on the western HB population as it has been assumed to be one large stable population (COSEWIC 2004a; 14  An Ecosystem Model of Hudson Bay, Canada    Luque and Ferguson 2010). The population of WHB belugas was set to 25000 whales (50000 whales at 50% of the time in the model area) t0 yield a biomass of 0.0247 t·km-2. In western Hudson Bay belugas feed on capelin (Mallotus villosus), river fish, marine worms and squids (Culik 2004), with capelin as an import contribution to the diet (Kelley et al. 2010). WHB belugas were assumed to feed on a slightly higher diversity of zooplankton due to the increased abundance found in WHB based on zooplankton samples (Harvey et al. 2006). The diet was set to 5% Arctic char, 2% Atlantic salmon, 15% gadiformes, 3% sculpins/zoarcids, 20% capelin, 1% sandlance, 4% other marine fish, 4% brackish fish, 5% cephalopods, 1% macro-zooplankton, 10%  euphausiids, 5% copepods, 10% crustaceans, 5% marine worms, and 10% other benthos. 3. Beluga James Bay It was assumed that the hunting on this population occurs primarily from Sanikiluaq as the whales hunted at this community have been shown to be different from the EHB belugas (de March and Postma 2003). Currently it is not fully known if this population is a separate population, or constant mixture of other populations, as they appear to be more closely genetically related to western HB belugas than eastern HB belugas (COSEWIC 2004a), although closer to eastern HB in proximity. Traditional knowledge indicates that there are some whales that spend the winter in James Bay, however it is not known if this is due to ice entrapment or not (Stewart and Lockhart 2005). Whales either remain overwinter in James Bay or migrate from the Quebec coast of HB into James Bay, with some migration around the Belcher Islands (Richard and Orr 2003 unpublished data as cited in Stewart and Lockhart 2005). Since 2004, eight belugas from James Bay have been fitted with satellite tags, and none have been shown to move into eastern HB (Hammill unpublished data cited in Gosselin et al. 2009).  For the model, the James Bay beluga will be treated as its own population, with hunting pressure occurring form the Sanikiluaq (Belcher Island) community, as no harvest occurs within James Bay (COSEWIC 2004a). Derived estimates of whale abundance have increased from roughly 1842 whales in 1985 to 3141 whales in 1993 to 7901 whales in 2001 (Gosselin et al. 2002). Estimates are considered conservative as they do not account for submerged animals, or those beyond survey view (Stewart and Lockhart 2005). This apparent increase in the population based on the 2001 survey is too high to be explained by population growth, and is believed to be an artefact of survey coverage, and seasonal movements (COSEWIC 2004a). A 2004 estimate of 3998 whales was believed to be too uncertain to use for management (Gosselin 2005). The model population was set to 1842 whales for the 1970s giving a biomass 0.00147 t·km-2. This estimate did not account for submerged animals, and should be doubled based on the correction factors of other beluga populations. However, assuming belugas spend 50% of their year in the model area, the abundance of 1842 was used as is for input.  The diet is believed to be focused heavily on capelin for this population (Stewart and Lockhart 2005) and was set to 1% Atlantic salmon, 5% gadiformes, 50% capelin, 5% cephalopods, 10%  euphausiids, 5% copepods, 10% crustaceans, 5% marine worms, and 9% other benthos.  Seabirds The group for birds includes all migratory and year round inhabitants. Most species arrive after the breakup of ice and leave before the freeze up, with a few exceptions of year round inhabitants (Stewart and Lockhart (2005). Some 133 species of birds are recorded to utilize the HB marine ecosystem (appendix 2) which funnels southbound migrating birds into James Bay, where the coastal marshes are an important stopover for many species (Stewart and Lockhart 2005).  Biomass for this group was estimated using bird counts from another Arctic area, the Chukchi Sea, Alaska, as Hudson Bay estimates were unavailable. The average number of birds from 1989-1991 in this region was 75 birds km-2 (Johnson et al. 1993). This coupled with the average weight of the bird species found with the Hudson bay area of 867g (Karpouzi 2005), gave a biomass estimate of 0.065 t·km-2. A P/B value of 0.113 year-1 was used for natural mortality, based on the seabird population in the Aleutian Islands (Heymans 2005), although a hunting mortality for HB based on catches of 0.005 year-1 was From the Tropics to the Poles, Wabnitz and Hoover  calculated. The combined P/B value of 0.118 year-1 was too low for the model and had to be increased to 0.37 year-1 in order to balance the model. The EE was set to 0.95, to let the model estimate Q/B.  Diet for this group, was based data provided by Karpouzi (2005), and was set to 2% seabirds, 3% Arctic char, 3% Atlantic salmon, 2% gadiformes, 3% sculpins/zoarcids, 15% capelin, 4% sandlance, 4% other marine fish, 10% brackish fish, 10% cephalopods, 12% macro-zooplankton, 5%  euphausiids, 1% copepods, 1% other meso-zooplankton, 2% marine worms, 3% echinoderms, 10% bivalves, 5% other benthos, 5% pelagic detritus.  Thick-billed murres have been monitored at Coats Island (in northern HB just southeast of Southampton Island) since 1985, and have shown an annual average increase in population (roughly 1.7% per year). Similar trends for thick-billed murres have been reported at Digges Island (just east of Coats Island at the northern edge of the model area) up until 2000 when the population appears to have levelled off (Gaston et al. 2009a). For the same region glaucous gulls have declined up to 50% (unpublished data cited in Gaston et al. 2009a). Near the Belcher Islands surveys show the mean number of gull nests declining by 50% since 1980 and slight declines of Arctic terns (only significant declines at 1 of 5 sites surveyed) (Gilchrist and Robertson 1999). The breeding of thick-billed murres has become earlier (6 days earlier since 1980), which is believed to be due to an earlier breakup of sea ice (17 days earlier when comparing 1988 to 2007), however it is not believed that changes to breeding cycles will be able to keep up with changes in environmental cycles (Gaston et al. 2009a; Gaston et al. 2009b). The diet of thick-billed murres has demonstrated shifts from Arctic cod to capelin (Gaston et al. 2003) (figure 4). Although local changes appear to have occurred, it is hard to extrapolate to all bird species from regional studies. No large scale increases or declines have been observed in HB that would apply to all bird species within this group, therefore no assumptions on trends has been made for this model.  Fish Fish species were determined based on the species named present in Hudson Bay and/or James Bay in appendix 3 of Stewart and Lockhart (2005). Species listed were categorized based on life history; marine, brackish, estuarine, diadromous, anadromous, or semi-anadromous. However, as the model is defined as the marine ecosystem only species listed as marine and some species defined as brackish were included in the Figure 4: Changes in diet of thick-billed murres at Coats and Diggs Islands from 1984-2002. Graph reproduced from data in Gaston et al. (2003). 01020304050607080901001984 1987 1990 1993 1996 1999 2002Contribution to Diet (%)YearArctic CodSculpins/ ZoarcidsCapelinSandlance16  An Ecosystem Model of Hudson Bay, Canada    model. There are ten groups of fish in the model, based primarily on familial traits and secondarily on life history characteristics. Species found in each functional group are listed in appendix 3.  As no comprehensive surveys have yet been completed, biomass was estimated for all fish groups, utilizing the ability of Ecopath with Ecosim to solve for one unknown parameter for each functional group. Biomass for all fish groups was estimated by the model using the inputs of P/B, Q/B, EE, and the diets of other functional groups. Total mortality was set to the sum of fishing mortality (table 2) and natural mortality, with the natural mortality being calculated using the life history tool page in Fishbase (Froese and Pauly 2008), which provides equation 6, where ܯ is the natural mortality, L∞= the maximum length of the fish, and ܶ is the temperature of the water in degrees Celsius (Pauly 1980; Froese and Pauly 2008).  As little information is known about fish in Hudson Bay, default values provide by Fishbase for L∞ were used. For temperature, both the average value provided for the species based on temperatures fish are normally found in (provided by Fishbase), and an average of 0.5˚C were used and calculated values are presented in table 3. The 0.5˚C value was chosen as it is the average water temperature for this region from 1960-2006, based on a global database of ice and sea surface temperature (SST) combining real and estimated data to obtain these values (Rayner et al. 2003; British Atmospheric Data Centre 2010).   (6)     ܯ ൌ 10ሺ଴.ହ଺଺ି଴.଻ଵ଼∗௅௢௚ሺ௅ಮሻା଴.଴ଶ∗்ሻ         Values for natural mortality (ܯ), Equation 6, was created using fish from tropical and temperate habitats and often underestimates mortality for polar species (Pauly 1980). Therefore, when considering all the species in group, higher values were generally chosen.  Fishing Mortality Fishing Mortality is likely to occur on all fish species in HB, as subsistence fishing is common. Catches from commercial fishery attempts have proven to be small and financially unsustainable, therefore there are no commercial fisheries operating in the model area at present, with only a few brief attempts in the past (Stewart and Lockhart 2005). Recreational fishery information is only available for Arctic char from 1988-1997 through DFO harvest records (DFO 1990; DFO 1991; DFO 1992; DFO 1993; DFO 1994; DFO 1995; DFO 1996; DFO 1997). Subsistence mortality was estimated using a per capita use rate derived from values provided by various sources from 1970-2001 (Anonymous 1979; Gamble 1988; Fabijian and Usher 2003) for the communities of Arviat, Paulatuk, and Inukjuaq as presented in Booth and Watts (2007). For fish, a per capita consumption rate of 30-120kg per person year-1 was estimated for 1970. Underreporting is believed to occur therefore the upper estimate of 120 kg per person year-1 was believed to be more accurate. Based on a population size of 10033 (see fishing section for community population estimates) this would yield a total catch of 1204.6 T of fish caught for subsistence hunting in 1970. This was divided among all fish groups except sharks and rays. Catch was divided among the different species groups based on sporadic community records of fish catches from 1975- 1990 as presented in table 14-8 of Stewart and Lockhart (2005). The contributions of total catches by species group are presented in table 2, and include the estimated hunting mortality.    Consumption rates were calculated using equation 7 from Palomares and Pauly (1998): Table 2. Fishing mortality based on per capita consumption rate of 120kg per person·year-1.  Species group % of total catch Catches (Tonnes) Hunting Mortality Arctic Char 35 421.614 6.334E-04 Atlantic Salmon 1 12.046 3.900E-05 Gadiformes 20 240.923 2.665E-04 Sculpins/ Zoarcids 20 240.923 6.604E-04 Capelin 10 120.461 1.823E-04 Sandlance 3 36.138 6.151E-05 Sharks/Rays 0 0 0 Other Marine Fish 5 60.231 9.137E-05 Brackish Fish 2 24.092 3.967E-04 From the Tropics to the Poles, Wabnitz and Hoover    (7)        logܳ/ܤ ൌ 7.964 െ 0.204݈݋݃ ஶܹ െ 1.965ܶᇱ ൅ 0.083ܣ ൅ 0.532݄ ൅ 0.398݀      where W∞ (infinity) is the weight a fish would reach if it grew to it L∞(the mean length of very old fish), T is the mean temperature in Kelvin, expressed as (1000 / (C + 273.15)) with C representing temperature in degrees Celsius. A is the aspect ratio of the caudal fin, h and d represent variables for feeding types; h=1 if the fish is herbivorous, h=0 if it consumes other food types, d=1 if the fish is a detritivore, d=0 if the fish consumes other food types. Again a temperature of 0.5ºC was used based on the average temperature for this region. The Ecotrophic Efficiency (EE) for all fish groups was set to 0.95 in order to allow the modelling program to estimate biomass parameters. Previous modelling indicates values close to one are widely used for mid-trophic level groups, indicating most of the organisms are consumed within the food web or from fishing, and relatively few die from old age (Christensen et al. 2005). The value 0.95 was chosen to assume 95% of the population will die from predation and fishing mortality (Christensen pers. Comm. 2006). Parameters calculated for all fish species are presented in table 3.  Arctic Char  The Arctic Char (Salvelinus alpinus) group consists of only one species. Char are anadromous, living primarily in marine waters (Stewart and Lockhart 2005). Due to the locations and increased availability for a short time period while in HB and JB, char are hunted by subsistence and recreational hunters (Stewart and Lockhart 2005). Arctic char in HB prey on amphipods, mysids, and fish (Stewart and Lockhart 2005). In Labrador the diet consists of fish (capelin, sand lance, and various sculpins), molluscs, crustaceans, insects, and chaetognaths (Dempson et al. 2002). Diet for the model was set to: 1% Atlantic salmon, 1% gadiformes, 1% sculpins/zoarcids, 2% capelin, 2% sandlance, 2% Other Marine Fish, 2% Brackish Fish, 10% macro-zooplankton, 5%  euphausiids, 31% copepods, 10% crustaceans, 10% other meso-zooplankton, 10% micro-zooplankton, 3% marine worms, 2% echinoderms, 3% other benthos, 4% primary production, 1% ice algae.  Atlantic Salmon The Atlantic salmon group also consists of only one species Salmo salar, which utilizes the marine environment during the winter in HB, JB, and HS. Although this species is not common in HB and JB it is harvested as bycatch, and is more prevalent in the Ungava Bay area just outside of the model area (Stewart and Lockhart 2005). Atlantic salmon is not known to be a major contributor to predator diets. Although region-specific studies have not been done, in other areas juveniles prey on a range of invertebrates (molluscs, crustaceans, and small fish), while adults have been known to prey on fish (capelin, sandlance, and small cod) (Froese and Pauly 2008). For the model the diet was set to: 1% Arctic char, 1% Atlantic salmon, 2% gadiformes, 2% sculpins/zoarcids, 5% capelin, 2% sandlance, 2% other marine fish, 3% brackish fish, 5% cephalopods, 15% macro-zooplankton, 8%  euphausiids, 8% copepods, 18% crustaceans, 3% other meso-zooplankton, 15% micro-zooplankton, 7% primary production, and 3% ice algae.  Gadiformes The Gadiformes group includes Arctic cod (Boreogadus saida), Greenland cod (Gadus ogac), and Polar cod (Arctogadus glacialis). These fish are important to the diets of many marine mammals in the area (see narwhal, ringed seal, harp seal, and beluga sections), although Arctic and Polar cod are more important to higher predators than Greenland cod. Arctic cod are believed to be declining, as their presence in the diet of thick-billed murres has declined since the 1980s (Gaston et al. 2003). Greenland cod in northern HB are omnivorous feeding primarily on benthic species; crabs, amphipods, polychaetes, and crustaceans, with few species consuming them, while Arctic cod take mostly copepods, hyperiid amphipods, ice-associated crustacea, and other pelagic prey, and are more important to higher predators than Greenland cod (Mikhail and Welch 1989). The diet for this group was set to 2% gadiformes, 5% capelin, 5% sandlance, 6% other marine fish, 3% crustaceans, 15% marine worms, 15% bivalves, 20% other benthos, 10% ice algae, and 4% ice detritus.  18  An Ecosystem Model of Hudson Bay, Canada    Table 3: Calculated input parameters for all fish groups within the model.  Group Species Common Name L∞ Average Temp oC Mortality at Average Temp M at 0.5 oC Q/B at 0.5 oC Arctic Char Salvelinus alpinus Arctic Char  1.5 0.10 0.10 1.7         Atlantic Salmon Salmo salar Atlantic Salmon 156 9 0.30 0.25 7.14         Gadiformes Arctogadus glacialis Polar cod 34 8 0.55 0.46 2.3  Boreogadus saida Arctic cod 31.3 1 0.31 0.30 2.5  Gadus ogac Greenland cod 79.5 1 0.22 0.22 1.3         Sculpins/ Zoarcids Gymnocanthus tricuspis Arctic staghorn 31.5 1 0.30 0.29 2.2  Icelus bicornis twohorn sculpin 16.6 1 0.51 0.50 3.6  Icelus spatula spatulate sculpin 22.1 3 0.35 0.33 3  Myoxocephalus quadricornis fourhorn sculpin 33.1 1 0.32 0.32 2.1  Myoxocephalus scorpioides Arctic sculpin 23.2 1 0.32 0.39 2.9  Myoxocephalus scorpius shorthorn sculpin 21.9 9.3 0.79 0.64 2.7  Triglops murrayi moustache sculpin 21.1 10 0.65 0.42 3.1  Triglops pingelli ribbed sculpin 27.3 10 0.35 0.28 3  Gymnelus viridis fish doctor 58.1 1 0.28 0.28 1.6  Lycodes pallidus pale eelpout 27.3 1 0.41 0.35 2.6  Lycodes reticulatus Arctic eelpout 37.6 1.3 0.30 0.28 2.2         Capelin Mallotus villosus capelin 16.9 4.3 0.85 0.78 3.9         Sandlance Ammodytes dubius northern sand lance 26.2 2 0.45 0.44 3.8  Ammodytes hexapterus stout sand lance 31.5 10 0.47 0.38 2.4         Sharks/Rays Somniosidae sleeper sharks   0.04  0.5  Rajidae skates   0.18  2         Other Marine Fish Leptagonus decagonus alligator poacher 22.1 1 0.45 0.41 3  Ulcina olriki Atlantic alligatorfish 9.2 1 1.03 0.77 5.3  Cyclopterus lumpus lumpfish 55 5 0.19 0.17 1.3  Eumicrotremus derjugini leatherfin lumpsucker NA    4.7  Eumicrotremus spinosus Atlantic spiny lumpsucker NA    4  Careproctus reinhardti sea tadpole 31.5 3 0.57 0.32 2.4  Liparis fabricii gelatinous snailfish 21.1 8 0.94 0.42 3.1  Liparis gibbus dusky snailfish 54 1 0.33 0.21 1.7  Liparis tunicatus kelp snailfish 16.9 1 0.98 0.49 3.5  Anisarchus medius stout eelblenny 31.5 1 0.26 0.32 2.4  Eumesogrammus praecisus fourline snakeblenny 23.2 1 0.35 0.39 2.9  Leptoclinus maculatus daubed shanny 21.1 1 0.38 0.42 3.1  Pholis fasciata banded gunnel 31.5 1 0.49 0.32 2.4  Clupea harengus Atlantic Herring 30.4 9 0.48 0.39 10.1         Brackish Fish Lumpenus fabricii slender eelblenny 38.1 1 0.28 0.28 2.2  Stichaeus punctatus Arctic shanny 14.5 12 0.94 0.55 3.9  Hippoglossoides platessoides Canadian plaice 70.4 1.4 0.19 0.18 1.7 NA indicates parameter could not be calculated due to missing information required for calculations.  From the Tropics to the Poles, Wabnitz and Hoover  Sculpins/Zoarcids Sculpins (Family: Cottidae) and zoarcids or eelpouts (Family: Zoarcidae) were combined to form one functional group and include: Arctic eelpout (Lycodes reticulates), Arctic sculpin (Myoxocephalus scorpioides), Arctic staghorn (Gymnocanthus tricuspis), fish doctor (Gymnelus viridis), fourhorn sculpin (Myoxocephalus quadricornis), moustache sculpin (Triglops murrayi), pale eelpout (Lycodes pallidus), ribbed sculpin (Triglops pingelli), shorthorn sculpin (Myoxocephalus scorpius), spatulate sculpin (Icelus spatula), and twohorn sculpin (Icelus bicornis). These two families were combined as nearly all members are small benthic fish found in shallow, mostly coastal waters. Of the eelpout species, only the fish doctor has been noted as important to predators, namely cods and sculpins, while the importance of pale and Arctic eelpouts are unknown (Stewart and Lockhart 2005). However, sculpins are consumed by cods, seabirds, seals, and other marine mammals, in addition to being caught for sport fishing occasionally (Stewart and Lockhart 2005).  The diets of these fish include plant materials, aquatic insects, crustaceans, benthic amphipods, polychaetes, bivalves, and detritus (Froese and Pauly 2008). The diet was set to 2% sculpins/zoarcids, 5% capelin, 5% sandlance, 4% other marine fish, 7% crustaceans, 15% marine worms, 11% echinoderms, 15% bivalves, 20% other benthos, 6% ice algae, and 10% ice detritus. Capelin Capelin (Mallotus villosus) is a marine species with a circumpolar distribution in the Arctic, sometimes occurring in brackish or freshwater, and is often found in schools (Froese and Pauly 2008). The population in HB is believed to be a surviving remainder from a warmer time period, likely the 1880s or earlier, with large swarms occurring in southern HB (Dunbar 1983). The ecology of adult capelin in HB is not well known (Stewart and Lockhart 2005), although they have been shown to be an important prey item to belugas, harp seals, and many bird species (Beck et al 1993; Gaston et al 2003; Loseto et al. 2009). Changes to the diets of thick-billed murres have identified a possible increase in capelin from 1980-2002 for birds located in the northern portion of HB (Gaston et al. 2003). The general diet of capelin is based on planktonic crustaceans, copepods, euphausiids, amphipods, marine worms, and small fishes (Froese and Pauly 2008). For the model the diet was set to 15% macro-zooplankton, 20% euphausiids, 20% copepods, 10% crustaceans, 5% other meso-zooplankton, 10% micro-zooplankton, 15% pelagic production, and 5% pelagic detritus.  Sandlance The sandlance group contains two species the northern sand lance (Ammodytes dubius) and the stout sand lance (Ammodytes hexapterus). Both species are small bottom dwelling fish which burrow in the sand and are important in the diets of forage fish, seabirds, and marine mammals (Stewart and Lockhart 2005). Sandlance feed on zooplankton, primarily copepods, crustaceans, and worms (Froese and Pauly 2008). The diet was set to 2% cephalopods, 5% macro-zooplankton, 15% euphausiids, 35% copepods, 5% crustaceans, 10% other meso-zooplankton, 15% micro-zooplankton 10% pelagic production, and 3% pelagic detritus.  Sharks/Rays  The Greenland shark (Somniosus microcephalus) and the thorny skate (Amblyraja radiate) are both bottom dwelling and likely very uncommon in HB and JB. The Greenland shark has been suggested to be present in HB, and the thorny skate is only noted to be found in James Bay, within the model area (Stewart and Lockhart 2005). Both are probably rare in the area, and not likely to be a significant contribution to fish biomass in general. Skates consume small fish and benthic invertebrates, while the Greenland shark consumes fish, seals, whales, and birds (Stewart and Lockhart 2005; Froese and Pauly 2008). The diet was set to 1% narwhal, 1% bearded seal, 1% ringed seal, 1% harp seal, 5% Arctic char, 2% Atlantic salmon, 15% gadiformes, 15% sculpins/zoarcids, 5% capelin, 8% sandlance, 1% sharks/ rays, 6% other marine fish, 4% brackish fish, 5% cephalopods, 5% macro-zooplankton, 1% euphausiids, 5% crustaceans, 5% marine worms, 10% echinoderms, 1% bivalves, and 3% other benthos.   20  An Ecosystem Model of Hudson Bay, Canada    Other Marine Fish The other marine fish group includes herring (family: Clupeidae), poachers (family: Agonidae), lumpfishs (family: Cyclopteridae), shannies (family: Stichaeidae), and gunnels (family: Pholidae), Species include: alligator poacher (Leptagonus decagonus), Atlantic alligatorfish (Ulcina olriki), Atlantic Herring (Clupea harengus), Atlantic spiny lumpsucker (Eumicrotremus spinosus), banded gunnel (Pholis fasciata), daubed shanny (Leptoclinus maculatus), dusky snailfish  (Liparis gibbus), fourline snakeblenny (Eumesogrammus praecisus), gelatinous snailfish (Liparis fabricii), kelp snailfish (Liparis tunicatus), leatherfin lumpsucker (Eumicrotremus derjugini), lumpfish (Cyclopterus lumpus), sea tadpole (Careproctus reinhardti), and stout eelblenny (Anisarchus medius).  These fish are all small benthic fish that live near varied substratum (mud, sand, and rocks), with the exception of herring, which are predominantly pelagic and schooling living from the surface to 200m. These fish are prey items for cod, seabirds, seals, other fish and lumpfish are noted to be eaten by Greenland sharks (Stewart and Lockhart 2005). Diets of these fish are focused on benthic and pelagic invertebrates, primarily crustaceans, polychaetes, clams, fish eggs, zooplankton, and herring have the ability to filter feed (Froese and Pauly 2008). The diet was set to 2% capelin, 1% cephalopods, 5% macro-zooplankton, 2%  euphausiids, 20% copepods, 20% crustaceans, 2% other meso-zooplankton, 5% micro-zooplankton, 6% marine worms, 5% bivalves, 5% other benthos, 10% pelagic production, 10% ice algae, and 7% pelagic detritus. Brackish Water Fish The brackish water group includes two species of shannies (family: Stichaeidae) which were considered to be brackish based; Arctic shanny (Stichaeus punctatus) and the slender eelblenny (Lumpenus fabricii) and one righteye flounder (family: Pleuronectidae), Canadian plaice (Hippoglossoides platessoides). Although all three of these species are found in inshore waters, they have been classified as brackish rather than marine and are consumed by larger marine fish and seabirds (Stewart and Lockhart 2005). The diets consist of invertebrates; crustacean, worms, and clams, in addition to small fish and fish eggs (Froese and Pauly 2008). The diet was set to 2% capelin, 2% sandlance, 2% brackish fish, 2% cephalopods, 17% macro-zooplankton, 5%  euphausiids, 5% copepods, 15% crustaceans, 5% other meso-zooplankton, 20% other meso-zooplankton, 2% marine worms, 2% echinoderms, 6% other benthos, 9% pelagic production, 1% ice algae, and 5% pelagic detritus.  Zooplankton Sampling of zooplankton has occurred twice in the HB region, once with a survey by Harvey et al. (2001) to sample the eastern side of HB in 1993, starting in JB and moving northward up the coast and into Hudson Strait. The second survey conducted in 2003 spanned from west to east just above 60˚N latitude (Harvey et al. 2006). Results from the surveys indicate higher zooplankton biomass on the western side compared to the eastern side of Hudson Bay, and increasing concentration as samples increased in latitude from James Bay up into Hudson Strait.  From the 1993 south to north survey (Harvey et al. 2001), biomass of samples ranged from 2.6 to 28.1 g· m-2. Original samples were presented in dry weight (0.52 to 5.62 g·m-2), but converted to wet weight using a conversion factor of 5 (DW:WW) for zooplankton (Cushing et al. 1958; Cauffopé and Heymans 2005). Samples were dominated by copepods, euphausiids, cnidarians, amphipods, and chaetognaths indicating sampling of the meso and macro-zooplankton (chaetognaths fall into the macro-zooplankton, while most other species are smaller and fall into the meso-zooplankton spectrum).  The 2003 east to west survey (Harvey et al. 2006) identified meso-zooplankton, dominated by copepods, to have 3 times more biomass than macro-zooplankton in Hudson Bay. This ratio was higher in Hudson Strait and Foxe Basin, up to 10 times more meso-zooplankton. Of the zooplankton standing stock 5-17% of the abundance of zooplankton sampled was macro-zooplankton for the HB portion with the chaetognaths Sagitta elegans as the most abundant. Wet weight of macro- and meso-zooplankton ranged from 5-10 g·m-2 for HB samples, although biomasses were higher for Hudson Strait, up to 20 g·m-2 for macro-zooplankton, and 110 g·m-2 for meso-zooplankton.  From the Tropics to the Poles, Wabnitz and Hoover  Cephalopods While little is known about cephalopods in HB, they appear in the diets of predators; birds, seals, and some whale species. Gonatus fabricii is an important prey item in the diets of thick-billed murres, and is the only species recorded within the model area (Gerdiner and Dick 2010). However, Rossia moelleri and other unidentified cephalopods have been recorded just outside the model area (Gerdiner and Dick 2010) indicating a strong possibility that more than one species is found within HB. This combined with the diets of predators led to the belief that cephalopods are present within the model area, and they were therefore included as a functional group. The biomass for cephalopods was estimated by the model given other parameters. The P/B and Q/B of 2.55 and 6.9 year-1 were taken from the cephalopod group in the 1979 Aleutian Island model (Heymans 2005). However, these values were adjusted in the balancing of the model to 1.5 and 5 year-1 for P/B and Q/B respectively. The EE for this group was set to 0.95. Diet for cephalopods was set to 1% Arctic char, 1% Atlantic salmon, 5% gadiformes, 5% sculpins/zoarcids, 8% capelin, 8% sandlance, 1% other marine fish, 4% cephalopods, 18% macro-zooplankton, 4%  euphausiids, 13% copepods, 10% crustaceans, 10% other meso-zooplankton, and 12% micro-zooplankton based on the diet of Antarctic cephalopods (Rodhouse and White 1995; Jackson et al. 2002).  Macro-Zooplankton The macro-zooplankton group includes all zooplankton species larger than 2mm. Chaetognaths (Sagita elegans) were the most abundant species from sample taken in eastern HB in 1993, with hydromedusa (Aeginopsis laurentii) being the second most abundant, and numerous unidentified species (Harvey et al. 2006). Biomasses from the 2003 survey were reported between 5-10 g·m-2. A value of 7.5 g·m-2 or t·km-2 was used for the biomass. P/B values of zooplankton larger than 1 mg WW for the Prince William Sound model ranged from 0.1 to 1.5 year-1  depending on the season, and Q/B ratios ranged from 0.33 to 5 year-1  (Okey and Pauly 1999). P/B for HB was set to 1 and Q/B set to 3 year-1 based on the values from Prince William Sound. Chaetognaths were the most abundant species in this group, with a diet focused on copepods (Tönnesson and Tiselius 2005). Other members of this group were believed to prey upon smaller zooplankton and phytoplankton species. The diet was set to 6.5%  euphausiids, 19% copepods, 2% crustaceans, 5% other meso-zooplankton, 30% micro-zooplankton, 22% pelagic production, 10.5% ice algae, and 5% pelagic detritus.  Euphausiids Euphausiids show increasing contribution to the meso-zooplankton biomass moving south to north (Harvey et al. 2001). Euphausiids consisted of Thysanhoessa rachii and other unidentified species. Based on the 1993 samples euphausiids contributed on average 2.14 g·m-2 or t·km-2 to the zooplankton biomass. The P/B for this group was set to 3 year-1 based on a krill larva value of 4, and adult krill value of 1 from the Antarctic Peninsula ecosystem model (Efran and Pitcher 2005). A P/Q ratio of 0.25 was assumed (Christensen et al. 2005), to allow the model to estimate both EE and Q/B. The diet was set to 1% macro-zooplankton, 0.1% euphausiids, 55.9% copepods, 1% crustaceans, 5% other meso-zooplankton, 10% micro-zooplankton, 15% pelagic production, 8% ice algae, and 4% pelagic detritus based on the diet of Antarctic euphausiids (Pakhomov et al. 1997; Cripps and Atkinson 2000; Atkinson et al. 2002). Copepods Small copepods dominate the meso-zooplankton biomass, up to 82% of total zooplankton biomass at on station in northern HB (Harvey et al. 2001). The average biomass over all stations sampled in 1993 was 4.015 g·m-2 or t·km-2, and thus was the biomass used for the model. Species include: Acartia longiremis, Calanus glacialis, Calanus finmarchicus, Calanus hyperboreus, Centropages hamatus, Metridia longa, and Pseudocalanus spp. as well as other unidentified species. P/B for the Prince William Sounds model copepod group was 5 year-1 (Okey and Pauly 1999). Other zooplankton groups show higher P/B values ranging from 5.8 to 36.3 year-1 for the Aleutian Islands (Heymans 2005) or 10.7 to 24 year-1  for the Kerguelen Islands (Pruvost et al. 2005). A P/B of 16 year-1 was used for the HB model. A P/Q of 0.25 was assumed to give a Q/B value of 64 year-1 when balancing the model. Copepods are primarily grazers, with 22  An Ecosystem Model of Hudson Bay, Canada    a strong link to ice algae identified in HB (Runge and Ingram 1987; Runge and Ingram 1991). Copepods have also been noted to consume other zooplankton species (Metz and Schnack-Schiel 1995). The diet was set to 5% micro-zooplankton, 70% pelagic production, 20% ice algae, and 5% pelagic detritus.   Crustaceans  The crustacean group includes all benthic crustaceans and zooplankton crustaceans (with the exception of euphausiids and copepods). The benthic and planktonic species were combined due to lack of distinction in the diet for higher predators. For the planktonic species this includes various Isopoda, Ostracoda, Amphipoda, Decapoda, and Cirripedia. Biomass for the planktonic component was averaged to 1.05 g·m-2 based on the 1993 survey. For the benthic component more species were identified (147 species compared to 5 identified for pelagic with many unknown) from Amphipoda, Cirripedia, Cumacea, Decapoda, Isopoda, Nebaliacea, Ostracoda, Pycnogonida, and Tanaidacea. In the Weddell Sea benthic Crustacea and Chelicerata contribute 0.45 g·m-2 or t·km-2 (Jarre-Teichmann et al. 1997). Although the contribution of benthic crustaceans is known in this area, it was estimated to be no more than the planktonic component. A biomass of 1.8 g·m-2 was used for the model. P/B for various crustacean plankton for Prince William Sound ranged from 2-8 year-1 (Okey and Pauly 1999). P/B for benthos ranged from 0.7 year-1 for benthic crustaceans in the Weddell Sea (Jarre-Teichmann et al. 1997) to 2.1 year-1 for benthic invertebrates for the Aleutian Islands (Heymans 2005). A P/B value of 3.6 year-1 was used along with a P/Q ratio of 0.25 to give a Q/B ratio of 14.4 year-1.  Antarctic amphipod diet consists primarily of detritus with some polychaetes, crustaceans, echinoderms and bryozoans (Dauby et al. 2001). In HB amphipods can significantly reduce the inshore algal biomass suggesting their ability to consume large amounts of producers (Stewart and Lockhart 2005). Benthic crustaceans were assumed to be primarily scavengers and carnivores. The diet was set to 1%  euphausiids, 5% copepods, 0.5% crustaceans, 1% other meso-zooplankton, 1% micro-zooplankton, 5% marine worms, 5% echinoderms, 5% bivalves, 10% other benthos, 30% pelagic production, 16.5% ice algae, 10% ice detritus, and 10% pelagic detritus.  Other Meso-Zooplankton The other meso-zooplankton group includes numerous unidentified species from the phyla Cnidarians, Annelida, Mollusca, and Urochordata. The average biomass for this group based on the 1993 survey was 1.21 g·m-2. The P/B was set to 10 year-1 based on overall zooplankton averages from the Prince William Sound model (Okey and Pauly 1999). The P/Q was set to 0.25 to give a Q/B of 40 year-1. Global analysis of meso-zooplankton consumption on primary producers indicated that in less productive marine systems meso-zooplankton were more reliant on alternative food sources such as protozoans and other zooplankton (Calbert 2001). For the HB region the diet was assumed to be 5%  euphausiids, 10% copepods, 2% crustaceans, 1% other meso-zooplankton, 10% micro-zooplankton, 45% pelagic production, 22% ice algae, and 5% pelagic detritus. Micro-Zooplankton The micro-zooplankton group includes all zooplankton smaller than 0.2mm. Sampling is not likely to include these smaller species as the mesh size in the nets is expected to let the smaller plankton through. Therefore there are no estimates of biomass for this group. Other model values for small zooplankton in the Aleutian Islands show a P/B ratio of 36 year-1 and a Q/B of 112 year-1 (Heymans 2005). Herbivorous zooplankton from the Kerguelen Islands were estimated to have a P/B of 24 year-1  and a Q/B of 96 year-1 (Pruvost et al. 2005). Okey and Pauly (1999) state a P/B of 15 year-1 for small zooplankton in Prince William Sound. For the HB model the P/B was set to the lower range of 15 year-1 and a Q/B of 45 year-1 was assumed. The EE for this group was set to 0.95. As micro-zooplankton are primarily grazers, although they have been noted to consume detritus in addition to ice algae in the winter months (Bathmann et al. 1993). The diet was set to 75% pelagic production, 17% ice algae, and 8% pelagic detritus.  From the Tropics to the Poles, Wabnitz and Hoover  Benthos There are few benthic species in the intertidal zone, however, below the sea ice the most common invertebrates are echinoderms, sea spiders, polychaetes, and worms (Stewart and Lockhart 2005). Various surveys of HB from 1953-1967 (Atkinsor and Wacasey 1989) identify presence of certain benthic species, however they fail to indicate abundance. From this survey there were 76 species of annelids identified, 157 arthropods, 53 cnidarians, 83 molluscs, 1 nemertean, 4 porifera, and 4 sipunculans. For each location species were recorded indicating which groups were present at the most locations. Benthos were split into four groups: marine worms, echinoderms, bivalves, and other benthos, primarily based on the diets of higher trophic level groups and their diets. Due to the lack of information for these species groups, parameters from other models of similar regions were incorporated and used for the benthic species. Parameter values of benthic invertebrates for other high latitude regions (Gulf of Alaska, Kerguelen Islands, and the Weddell Sea) are presented in table 4. Of the models built for higher latitudes, the Weddell Sea model is most comparable to the HB region, as the Gulf of Alaska and Kerguelen Islands are more open, productive ecosystems, while the Weddell Sea has less mixing compared to the other two. Brey and Gerdes (1998) found community P/B ratio to increase from 0.18 to 0.55 year-1 as depth increases for the Weddell and Lazarev Seas (Antarctica). For all benthic groups biomass was estimated, using inputs for P/B, Q/B, and a value of 0.95 for the ecotrophic efficiency.  Marine Worms The marine worm functional group includes all phyla of worms; Nematoda (round worms), Phoronida (horseshoe worms), Priapulida (priapulid or penis worms), Sipuncula (peanut worms), and Annelida (bristle worms). P/B and Q/B values of 0.6 and 4 year-1 respectively, were used based on the Weddell Sea model for the group “polychaetes and other worms” (Jarre-Teichmann et al. 1997), along with an EE of 0.95. Feeding types range from deposit feeders (Polychaetes) to trap feeders (Sipunculans) (Por and Bromley 1974; Brock and Miller 1999). The diet was set to 1% macro-zooplankton, 1%  euphausiids, 3% copepods, 1% crustaceans, 2% other meso-zooplankton, 3% micro-zooplankton, 1% marine worms, 1% echinoderms, 10% other benthos, 4% pelagic production, 12% ice algae, and 61% ice detritus.  Echinoderms The echinoderm functional group contains all species under the phylum Echinodermata, which includes the following classes: Asteroidea (Sea stars), Crinoidea (sea lilies), Echinoidea (sea urchins), Holothuroidea (sea cucumbers), and Ophuiroidea (brittle stars). The P/B and Q/B ratios were taken from all echinoderm groups in the Weddell Sea model and averaged to give 0.164 and 0.63 year-1 (Jarre-Teichmann et al. 1997). However, these values were too low to balance the model, so they were increased to 0.3 and 1 year-1  (P/B and Q/B) to the higher limits of this phylum for the Weddell Sea model (as the values for Crinoidea) to balance the model. The diet was set to 1%  euphausiids, 2% copepods, 5% crustaceans, 1% other meso-zooplankton, 3% micro-zooplankton, 10% marine worms, 1% echinoderms, 10% bivalves, 15% other benthos, 3% pelagic production, 8% ice algae, and 41% ice detritus to account for a range of feeding modes. Sessile echinoderms rely on suspended particles, while more active echinoderms such as seastars are able to actively hunt prey and most likely feed on other benthic species in the region.  Bivalves HB bivalves are from the class Pelecypoda (phylum Mollusca). This class was given its own functional group due to its importance to walrus, bearded seals, and fish. The P/B and Q/B values of 0.57 and 6.33 year-1 were taken from the Newfoundland model (Heymans 2003) and used for the HB values. The EE was set to 0.95 and the biomass was estimated by the model. As suspension feeders, bivalves were assumed to prey on species likely to come in contact with them. The diet was set to 3% copepods, 5% other meso-zooplankton, 5% micro-zooplankton, 5% pelagic production, 12% ice algae, and 70% ice detritus.   24  An Ecosystem Model of Hudson Bay, Canada     Other Benthos  The other benthos group includes all other invertebrate species found within HB. Those which have been named by Atkinsor and Wacasey (1989) include molluscs (Scaphopods or tusk shells), porifera (sponges), Pycnogonida (Arthropod: sea spiders), Ascidiacea (sea squirts), Brachiopoda (lamp shells), Cnidarians; anthozoa and hydrozoa (anemones/ corals and hydroids), Bryozoa (moss animals). Based on the benthic invertebrate groups from the Gulf of Alaska (Heymans 2005), the shallow benthic omnivores from the Kerguelen Islands (Pruvost et al. 2005), and the other benthic invertebrates from Newfoundland (Heymans 2003), the P/B was set to 2.5 year-1, and the Q/B was set to 12.5 year-1. The EE was set to 0.95 and the biomass was estimated for this group. A general diet was set to 1% macro-zooplankton, 1% other meso-zooplankton, 1% micro-zooplankton, 1% marine worms, 1% echinoderms, 1% bivalves, 1% other benthos, 5% pelagic production, 22% ice algae, and 66% ice detritus, as there are a variety of feeding types in this group. Table 4: Parameters for benthic functional groups from high latitude Ecopath models Functional Group Model Area Year of Model B  (t·km-2 ) P/B (year) Q/B (year) Reference Epibenthic Carnivores Gulf of Alaska 1963 35.601 2 17 (Heymans 2005) Benthic Invertebrates Gulf of Alaska 1963 5.194 0.98 6.553 (Heymans 2005) Deep benthic omnivores Kerguelen Is. 1987 30 3 10 (Pruvost et al. 2005) Shallow benthic omnivores Kerguelen Is. 1987 3.1 2.1 10 (Pruvost et al. 2005) Shallow benthic carnivores Kerguelen Is. 1987 8.7 2 10 (Pruvost et al. 2005) benthic mollusca Weddell Sea 1980s NA 0.3 1 (Jarre-Teichmann et al. 1997) Tunicata Weddell Sea 1980s 2.8 0.3 1 (Jarre-Teichmann et al. 1997) Porifera Weddell Sea 1980s 4.81 0.18 0.6 (Jarre-Teichmann et al. 1997) Hemichordata Weddell Sea 1980s 6.26 0.3 2 (Jarre-Teichmann et al. 1997) Lophophora and Cnidaria Weddell Sea 1980s 7.49 0.1 1 (Jarre-Teichmann et al. 1997) Benthic Crustacea and Chelicerata Weddell Sea 1980s 0.45 0.7 3.5 (Jarre-Teichmann et al. 1997) Polychaeta and other worms Weddell Sea 1980s 27.51 0.6 4 (Jarre-Teichmann et al. 1997) Echinoidea Weddell Sea 1980s 0.54 0.07 0.233 (Jarre-Teichmann et al. 1997) Crinoidea Weddell Sea 1980s 6.2 0.3 1 (Jarre-Teichmann et al. 1997) Ophiuroidea Weddell Sea 1980s 24 0.173 0.577 (Jarre-Teichmann et al. 1997) Asteroidea Weddell Sea 1980s 20.88 0.08 0.267 (Jarre-Teichmann et al. 1997) Holothuroidea Weddell Sea 1980s NA 0.2 1.1 (Jarre-Teichmann et al. 1997) Large Crabs Newfoundland 1995-1997 0.232 0.3 1.2 (Heymans 2003) Small Crabs Newfoundland 1995-1997 1.942 0.3 1.5 (Heymans 2003) Lobster Newfoundland 1995-1997 0.003 0.38 4.42 (Heymans 2003) Shrimp Newfoundland 1995-1997 1.859 1.45 9.667 (Heymans 2003) Echinoderms Newfoundland 1995-1997 112.3 0.6 6.667 (Heymans 2003) Polychaetes Newfoundland 1995-1997 10.5 2 22.222 (Heymans 2003) Bivalves Newfoundland 1995-1997 42.1 0.57 6.333 (Heymans 2003) Other Benthic Invertebrates Newfoundland 1995-1997 7.8 2.5 12.5 (Heymans 2003) From the Tropics to the Poles, Wabnitz and Hoover  Primary production Primary production in the model was split into two groups; pelagic production and ice algae. Pelagic production refers to the producers which bloom in the springtime in a seasonal pulse and are not generally available to the food web the remainder of the year. The ice algae group represents the species which are frozen into the sea ice in the fall and are released when the sea ice melts. Many of the species frozen within the ice are accessible throughout the winter via brine channels in the ice.  Numerous species of producers exist including: dinoflagellates, Prasinophytes, cryptophytes, chryophytes, centric diatoms, chlorophytes, flagellates, Prymnesiophytes, and pennate diatoms (Harvey et al. 1997). Two surveys of phytoplankton have been completed in HB; one in 1993 sampling from James Bay up the east coast of Hudson Bay into Hudson Strait (Harvey et al. 1997), and a second in 2003 running east to west through the middle of HB (Harvey et al. 1997). The first survey in 1993 yielded estimates of 0.36-133.5 t·km-2 (based on chl a samples of 1.2-145 mg·m-2)2, and the second survey estimated 7.5-75 t·km-2 (based on chla samples of 25-250 mg·m-2). Pelagic Production Pelagic production was sampled at 0.33-129 t·km-2 (1.1-431 mg Chla m-2) in 1993 (Harvey et al. 1997), although this was during the ice free season, so a biomass of 8 t·km-2 was assumed as the starting value. The EE was set to 0.8 to represent a 20% sinking rate to detritus, and the P/B ratio was estimated by the model.  Ice Algae The ice algae contribution to primary production was sampled to be 0.03-4.2 t·km-2 (0.1-14 mg Chla m-2) in 1993 (Harvey et al. 1997), 0.003-6 t·km-2 for values ranging 1978-1990 (Legendre et al. 1996), and 0.03-3.6 t·km-2 in 1986 (Tremblay et al. 1989). The contribution is thought to be slightly higher at the start of the model in 1970, as the extent of sea ice has decreased since this time. Ice algal contribution to total production has been estimated at 25% in Hudson Bay (Legendre et al. 1996) and ranging from to 57% of all production in the central Arctic to 3% in surrounding sub-Arctic areas (Gosselin et al. 1997).  Biomass of algae within the ice has reached levels of 0.6gC m-2 in the Antarctic (Weddell Sea and Antarctic Peninsula) during spring and fall in the 1980s (Garrison and Buck 1989), giving a biomass of 5.4 t·km-2 (Pauly and Christensen 1995)3. The biomass for HB was set to 3.5 t·km-2.  The EE for ice algae was set to 0.65, to account for the export of producers from the ice algae to ice detritus. Based on Tremblay et al. (1989), at least 20% of ice algal production during the spring was exported to the benthos, with 30% remaining in the pelagic zone, and another 50% thought to remain in the water column. As a yearly average, it was assumed that 45% of ice algae was exported to the ice detritus group, resulting in an EE of 0.65. The P/B was estimated by the model.  Detritus Detrital biomasses was calculated using equation 8 (Pauly et al. 1993):  ሺ8ሻ																				Logଵ଴D ൌ െ2.41 ൅ 0.954	Logଵ଴PP ൅ 0.863Logଵ଴10E where D is the standing stock of detritus (g C m-2 year-1), PP is primary productivity (g C m-2 year-1), and E is the euphotic depth (in meters).                                                    2 Wet weight (t·km-2) was calculated using the conversion Chla=1.5% of ash free dry weight (AFDW) (Farabee 2001), 1g carbon=2g AFDW (Cauffopé and Heymans 2005), and 1gC=9g wet weight (Pauly and Christensen 1995). 3 Using the conversion for phytoplankton where 1g Carbon=9g wet weight.  26  An Ecosystem Model of Hudson Bay, Canada    Ice Detritus In April the maximum ice thickness is 1.5m with over 85% of HB being covered in sea ice (Danielson 1971). To calculate ice detritus an average euphotic depth of ice algae was assume to be 0.5m, combined with the ice algae biomass gave an ice detritus biomass of 0.009 t·km-2. Pelagic Detritus  For the pelagic detritus group, a euphotic depth of 50m was used (Harvey et al. 1997), to give a value of 0.33 t·km-2.  Ecosim Parameters Fisheries In order to incorporate hunting and fishing pressure on various species, numerous “fisheries” were created within the model to account for catches within the first year (1970), which was then continued through the temporal simulations in Ecosim. Catches for the first year, and subsequent years are presented.  Polar bear Hunting 1. Western Hudson Bay Polar Bears The average catch for the 1980s for WHB bears was 44 (Lee and Taylor 1994), and then increased to an average of 46.8 bears from 1999-2004 (Aars et al. 2005). Catches were set to 44 from 1970-1998, 46.8 from 1999, and then 47 from 2005-2010 based on the 2005 quota of 47 (Aars et al. 2005). Initial catch for 1970 was set to 44 bears.  2. Southern Hudson Bay Polar Bears  The average catch of SHB polar bears for the 1980s was 68 (Lee and Taylor 1994) and decreased to an average of 40.4 from 1999-2004 (Aars et al. 2005). Catch was assumed to be 68 bears per year from 1970-1990, and then decreased to 40.4 from 1991-2004. The annual quota in 2005 was set to 25 bears (Aars et al. 2005). Catches from 2005-2010 were set to 25 bears. For 1970 the catch was set to 68 bears. 3. Foxe Basin Polar Bears Average catches for the 1980s were 142 bears (Lee and Taylor 1994) and decreased to an average of 97.4 for 1999-2004 (Aars et al. 2005). Catches were assumed to be 142 bears from 1970 to 1990, and then 97.4 bears each year until 2005, where the quota was raised from 97 to 106. It was assumed 106 bears were harvested each year from 2005-2010. For modelling purposes, these values were reduced to 20% to reflect the adjustments in biomass regarding the population size within the model area, as 20% of the Foxe Basin population resides in the model area. For 1970 the catch was set to 28.4 bears.  Killer Whale Hunting  Killer whales are not generally targeted, however they are occasionally hunted in HB (Ferguson pers. comm.). Table 5 identifies known harvests of killer whales in the eastern Canadian Arctic (Higdon 2007). These values were used for the HB killer whale population as relative catches. For 1970, there were no reported catches; however a value was needed in the model. The equivalent biomass of ¼ of a whale was used as a starting value.   Table 5: Known killer whales harvests in the eastern Canadian Arctic (Higdon 2007) Year Number of Whales Harvested 1978 1 1981 12 1995 1 2000 5 From the Tropics to the Poles, Wabnitz and Hoover  Narwhal Hunting In general narwhals are hunted during their migration and through the summer months primarily by Repulse Bay, with some involvement from other communities; Chesterfiled Inlet, Coral Harbour, Ranklin Inlet, Whale Cove, and Cape Dorset (DFO 1998; Westdal et al. 2010). The annual quota for the communities within HB is currently listed at 112 whales per year.   Catches of narwhal by Repulse Bay are shown in figure 5, not including the struck and loss rate. Catches for 1970 were set to 6 whales, the same value for 1978, which is the first year there were any recorded catches. Reported struck and loss rates range from of 40% of total catch (Roberge and Dunn 1990), to 12-56%, with specific hunts up to 71% (Weaver and Walker 1988) as observed by the Department of Fisheries and Oceans (DFO). However, non-DFO observers of the hunt have commented on how hunters only take sure shots when being officially observed (Nicklen 2007), meaning more whales are likely struck than the DFO statistics imply. The struck and loss term generally only accounts for whales known to die. Superficially wounded whales are not included in these estimates, even though they may not survive. Records also do not account for unreported catches.  Although the biomass was adjusted to 50% to account for half of the year (and feeding) to occur within the model area, as the catch data excludes the struck and loss rates, and underreporting. Catches were taken as is, without adjusting for the reduced biomass in the model, indicating mortality from catches is double than reported in figure 5.    Bowhead Hunting There have been 6 recorded kills of bowhead whales from the HB region; 1994 (unlicensed- Foxe Basin), 1996 (Repulse Bay), 1998 (Cumberland Sound), 2000 (Coral Harbour) 2003 (northern Foxe Basin), and 2005 (Repulse Bay) (Higdon 2008). From 1918-1988 Inuit from Greenland and Canada killed an estimated 36 bowhead for harvest and another 14 were struck and lost (Higdon 2008) since the end of commercial whaling, meaning of the 50 whales killed only 72% were harvested. In addition to the 6 recorded kills, a struck and loss rate of 25% was assumed (Ferguson pers. comm.) from 1994 onwards, meaning roughly 1 whale was killed every four years in addition to the 6 recorded kills. The catch for 1970 was set to 1 whale, with no catches until 1994 in the model. 040801201602001977 1982 1987 1992 1997 2002 2007Narwhals HarvestedYearFigure 5: Reported catches of narwhal from 1977-2007 for Repulse Bay, Chesterfiled Inlet, Coral Harbour, Ranklin Inlet, Whale Cove, and Cape Dorset (DFO 1990; DFO 1991; DFO 1992; DFO 1993; DFO 1994; DFO 1995; DFO 1996; DFO 1997; Stewart and Lockhart 2005). Figure does not incorporate a struck and loss rate.  28  An Ecosystem Model of Hudson Bay, Canada    Walrus Hunting  Hunting of walrus has been estimated at 35+ animals for south HB walrus and 230 for NHB walrus each year (NAMMCO 2005a). However, reported landings are less than half of these estimated values. Hunting for southern walrus occurs in Sanikiluaq, Kuujjuarapik, Umiujaq, and Inukjuak while the Northern Walrus group incurs hunting pressure from Whale Cove, Rankin Inlet, Chesterfield Inlet, Repulse Bay, and Coral Harbour. Catches from 1972-1987 (Strong 1989) and 1993-2003 from multiple sources summarized in (Stewart and Lockhart 2005) were used to fit the model. Discrepancies between the two data sets stem from coverage of different communities.  1. Southern Walrus  For the southern walrus population, the 1972-1987 dataset only includes Sanikiluaq, were the 1993-2003 dataset also includes the communities Kuujjuarapik, Umiujaq, and Inukjuak, which almost certainly had catches for the earlier time period. The inclusion of more communities from 1993-2003 may artificially inflate hunting pressure within the model. However, despite the lack of more inclusive data from 1971-1987, the data is used “as is” and is used as relative catches to fit the model. Catches from 2004 onwards were set to the 2003 reported landings. The 1970 catch was set to 8 animals, the same as the catch in 1972, as there were no records of catches for 1970.  2. Northern Walrus For the northern walrus population, the same 1971-1987 dataset includes catches from Whale Cove, Rankin Inlet, Chesterfield Inlet, Repulse Bay, and Coral Harbour. The latter dataset also includes Arviat, Ivujivik, Akulivik, and Puvirnituq. Again the catches were used as relative catches to fit the model. Catches from 2004 onwards were set to the reported value to 2003. Starting value for 1970 was 74 walrus, the same as the 1972 landings. Beluga Hunting All populations of beluga are hunted; however catch statistics do not distinguish between the stocks. Catches from western HB communities (Baker Lake, Chesterfield Inlet, Coral Harbour, Rankin Inlet, Repulse Bay, Sanikiluaq, and Whale Cove) were presumed to harvest the WHB beluga stock due to proximity. EHB and JB belugas are landed from communities on the eastern side of HB;  Kuujjuarapik, Umiujaq, Inukjuak, Puvirnituk, Akulivik, and Ivujivik from Nunavik, and Sanikiluaq from Nunavut, as both groups migrate down the eastern coast of HB to their summering locations. Of the whales landed in Sanikiluaq (Belcher Islands), it was assumed that half were from the JB beluga group, and half were from the EHB beluga group, as tagging studies show eastern HB and JB belugas located around the Belcher Islands (de March and Postma 2003). For the communities along the eastern coast of HB (Nunavik), the catches were thought to be mostly (70% of catches) from the EHB belugas, as the belugas not only using this as a migration route, but also summering in these areas. JB belugas use the same migration path, but move through to the summering location in James Bay making them available 0501001502002503001970 1975 1980 1985 1990 1995 2000 2005Catches (# of Whales)YearEHB BelugaJames BelugaWHB Beluga      Figure 6: Catches of Beluga whales form 1970-2007 as aggregated by stock.  From the Tropics to the Poles, Wabnitz and Hoover  to hunters for a shorter period of time. The remaining 30% of catches from the Nunavik communities was determined to be from the JB beluga group. Catches from 2008-2010 were set to the 2007 value, for all groups. Figure 6 identifies the trends in beluga harvest rates from 1970-2007 by stock, with landings per community taken from the Joint Commission on narwhal and beluga data (JCNB 2009). As in the case with narwhals, a struck and loss rate was incorporated. Reports on 3 communities (1 within the model area) indicates mortality nearly 10 times higher than reported catches when struck and loss rates are considered. When considering loss rates from narwhal, this value appears high. However, as the biomass was adjusted to 50% to account for time within the model area, the catches were not. This assumes double the hunting mortality on all beluga stocks than is reported.   1. Beluga East Catches for 1970 were set to the 1974 value of 83 whales, based on catches (JCNB 2009), and delineation of catches per community. 2. Beluga West Catches for 1970 were set to the 1976 value of 152 whales. 3. Beluga James Catches for 1970 were set to the 1974 value of 35 whales. Sealing, Bird Hunting, and Fishing In some cases catches were inferred based on a per capita basis, for many unregulated species. In these instances the increase in human population is used to calculate an increase in catches. Human community population size was used to estimate the harvest of birds, seals and fish. The human population in the Nunavut portion of Hudson Bay has more than doubled from 1981-2006, increasing from 4686 to 9491 inhabitants (Statistics Canada 2006); however estimates before this are not available.  Using the data from 1981 to 2006, a linear regression was fit to the data to estimate the growth rate giving an R2 value= 0.996 (figure 7). The growth pattern was assumed to decline constant from the 1970-1981 time period, lacking better data. As community data for Nunavik was not as readily available, population growth was presumed to follow the same growth pattern as communities in Nunavut. In 2006 the total human population for all communities in HB (Nunavut and Nunavik) was 30,117 (Bell 2002; Statistics Canada 2006; Nunavut Bureau of Statistics 2008; Sutherland et al. 2010). Following a linear decline in growth rate (figure 7), this estimated the population to be 10,033 individuals for all Hudson Bay communities in 1970. This value was used to calculate hunting rates for seals, birds, and fish. 051015202530351975 1980 1985 1990 1995 2000 2005 2010Nunavut Population Size (x1000)YearFigure 7: Regression of community population size in Nunavut (all communities) from 1981-2006. 30  An Ecosystem Model of Hudson Bay, Canada    1. Sealing Seal hunting is not currently regulated, although some estimates have been collected by community for 1975-1985 as summarized in table 14-9 various sources (Stewart and Lockhart 2005). Based on the number of seals caught, species, and community population, a per capita hunting rate of 1.1 seals per person year-1 was used. Catches were broken down based on the number of each seal species killed, resulting in 92.6% ringed seals, 6.1% bearded seals, 1% harp seals, and 0.3% harbour seals. The total number of seals caught in 1970 was set to 9110. Number of people was used to drive effort of seal catches, with the proportion of each seal species remaining constant.  2. Bird Hunting Hunting of birds is not regulated and Inuit do not require a license. Birds, eggs, down and other inedible products can be harvested any time of the year by Cree or Inuit (Migratory Birds Convention Act 1994). Based on survey records of bird harvests per community during 1975-1985 from table 14-10 (Stewart and Lockhart 2005), it was estimated that an average of 21.3 birds were harvested for every member of the community. The catches for 1970 were set to 213,703 birds.  3. Fishing Fishing rates were based on a per capita rate of 120kg per person year-1. See Fishing mortality (table 2) for breakdown of catches. Catches for 1970 were set to 1204 tonnes, with effort being driven by the number of people in the community.   Fitting the Model to Data Time series data (table 6) was read in as catches or abundance trends. For unregulated fisheries or hunting activities based on the size of the human population (fishing, bird hunting, and sealing), effort was driven by human population size4 (figure 7).  Forcing Functions The model is based on an understanding of the effects of climate change on the ecosystem. Warmer air temperatures, caused by climate change, have altered the mean ice freeze-up and break-up dates by 0.8-1.6 weeks in spring and fall (Hochheim et al. 2010). Figure 8 uses data from the HadISST5 (Hadley Centre Sea Ice and Sea Surface Temperature data set) model (British Atmospheric Data Centre 2010) to show the average % cover of sea ice for HB by month, with 95% CI. Starting in June, the variation in average ice cover increases, with June, July, November, and December having the greatest variance in ice cover. The SST also                                                  4 Human population was scaled to 1 for 1970. 5 The HasISST dataset has replaced the Global sea Ice and Sea Surface Temperature (GISST) dataset. HasISST data is generated from other datasets in addition to sea ice algorithm, and passive satellite data http://hadobs.metoffice.com/hadisst/    Table 6:  Name and type of time series data used to fit the HB model Time Series Data Type of Data Bowhead Abundance Relative Abundance Bowhead Catches Forced Catches Foxe Basin Polar Bear Abundance Relative Abundance Foxe Basin Polar Bear Catches Relative Catches Western HB Polar Bear Abundance Relative Abundance Western HB Polar Bear Catches Forced Catches Southern HB Polar Bear Catches Relative Abundance Narwhal Catches Forced Catches Eastern HB Beluga Abundance Relative Abundance Eastern HB Beluga Catches Forced Catches Western HB Beluga Abundance Relative Abundance Western HB Beluga Catches Forced Catches James Bay Beluga Abundance Relative Abundance James Bay Beluga Catches Forced Catches Northern HB Walrus Catches Forced Catches Southern HB Walrus Catches Forced Catches Killer Whale Abundance Forced Abundance Killer Whale Catches Forced Catches Arctic Cod Abundance Relative Abundance Sculpin/Zoarcid Abundance Relative Abundance Capelin Abundance Relative Abundance Sandlance Abundance Relative Abundance   From the Tropics to the Poles, Wabnitz and Hoover  becomes increasingly variable from June to December, and it is these changes in temperature and ice freeze-up and break-up dates that are thought to be important driver in the ecosystem and hence are implemented in the model. The availability of ice algae within the model is contingent upon the presence of sea ice; therefore the ice algae group was driven through a forcing function (FF) in the model. The sea ice FF was applied to the ice algae group, as a multiplier of the production rate using the average % cover of sea ice of all cells in the model area. The data was rescaled to all positive values with a mean value of 1 for the first year (1970). The pelagic production functional group was also driven in the model through SST, using the same HadISST dataset. Figure 8 shows the annual SST average for HB by month, with 95% CI. Again, data was rescaled to positive values with a mean value of 1 for 1970.     Mediation Functions In order to fit the polar bear groups (FB, WHB, SH), a mediation function was used. Sea ice is critical to polar bear foraging, as they use the ice as a hunting surface (Stirling and Derocher 1993). Declines in the western HB polar bear population from 1981-1998 have been linked to earlier breakup of the ice in the spring, and has been shown to cause reproductive stress and decreased body condition (Stirling et al. 1999). These effects have only been shown to be significant for the western HB population, as the timing of sea ice break-up has changed only on the west coast of HB (Stirling et al. 1999; Stirling and Parkinson 2006). A mediation function was applied to all polar bear groups, based on the changes in western HB. A sigmoid shape function was used with ice algae as the mediating group (figure 9). As the biomass of ice algae increases (which is driven by the % sea ice cover, making it a proxy for sea ice), polar bears have a larger foraging area and their prey becomes more vulnerable to them. For the starting point, near the top of the curve was selected, as changes in the sea ice have been documented locally since the 1980s (Gaston et al. 2009b). The sigmoid shape was selected, as it is believed once the sea ice reaches a maximum/minimum, there is no added benefit/detriment to polar bears. The reference point on the Figure 8: Monthly SST and ice cover with 95% CI for 1970-2009. Values were taken from the HadISST model.  32  An Ecosystem Model of Hudson Bay, Canada    curve which crosses the y-axis at 1 indicates the 1970 or starting value of the model, meaning increases in sea ice will have smaller effects on polar bears than decreases in sea ice.  Although declines in the Foxe Basin and southern HB populations of polar bears are not believed to be as extreme as WHB, it is highly likely they will respond to declines in sea ice the same way. Therefore the same mediation function was applied to all polar bear functional groups. Biomass Accumulation Abundance of Eastern HB belugas has declined from 1985-2008 (Hammill 2001; Hammill et al 2009b), however the model was unable to capture this decreasing trend through hunting and predation alone. Moreover, the model was unable to capture the large increases in the JB beluga population. As the JB stock of belugas is genetically different from EHB belugas, it is hypothesized this stock is a constant mixture of other stocks (de March and Postma 2003). Migration from EHB belugas to JB belugas was incorporated into the model in the form of biomass accumulation to assist in fitting. A decrease of 0.5% year-1 was necessary to fit the observed declines of EHB belugas. This led to an increased biomass accumulation of 1% year-1 to JB belugas, as the biomass of this group was roughly half the EHB biomass. Both P/B values were adjusted to accommodate for these changes; EHB beluga P/B was decreased from 0.0758 to 0.0658 year-1, and JB beluga P/B was increased from 0.0673 to 0.0873 year-1. A positive biomass accumulation rate was also used for bowhead whales, as the population is still rebounding from heavy commercial harvests (Higdon 2008 unpublished data), and the increases were not able to be captured by the model. A rate of 2% year-1 was initially used, however this value was later lowered to 0.7% year-1, and was still able to capture the increase.  Group Info Parameters The default maximum relative feeding time default of 2 was used for all species except marine mammals where it was set to 10 for all whale species (killer, narwhal, bowhead, and belugas), and 5 for all pinniped groups (walrus, harp, ringed, bearded, and harbour seals). The feeding time adjustment rate default of 0 was used for all species groups except marine mammals where it was set to 0.5 (Christensen et al. 2005; Christensen et al. 2007). Vulnerabilities Vulnerabilities were first estimated using the automated fit to time series routine in Ecosim (Buszowski et al. 2007). Next, the vulnerabilities for individual predator prey interactions were adjusted to fit the model more accurately to time series data. All vulnerabilities are displayed in appendix 4. Figure 9: Polar Bear Mediating Function with ice algae as the mediating group (x- axis). Y axis shows the relative weight of polar bears, starting at y=1 (Ecopath value).  From the Tropics to the Poles, Wabnitz and Hoover  RESULTS Balancing the Model Many parameters were refined during the balancing process, through a series of steps. A general outline of the progression is presented, although adjustments to the diets were also made but not noted. Final parameter values of the balanced model are presented in table 7.  After creating all the functional groups and calculating general parameters, and diets, fishing groups were created. Once the catches for 1970 were determined, P/B ratios were adjusted to include hunting and fishing mortalities. After adjusting the P/B for marine mammals, birds, and fish, the P/B of fish had to be increased further.  The equation used to calculate P/B for fish often underestimates higher latitude species (Pauly 1980), and the smaller P/B was causing the model to estimate large biomasses of fish. Consequently, these ratios were increased to the upper limits based on the species found within the functional group. Many of the zooplankton groups lacked region specific data for P/B and Q/B, therefore a P/Q ratio of 0.25 was assumed, so the model could estimate an additional parameter.  The EE of birds was too high indicating too much mortality. The P/B ratio was increased to allow enough hunting and predation mortality to occur in the model. Impacts of each functional group upon others are presented in appendix 5, as output from the mixed trophic impact table in Ecopath.   34  An Ecosystem Model of Hudson Bay, Canada     Table 7: Balanced model with parameters estimated by the model in bold Group Name Trophic Level Biomass (t· Km-2) P/B (Year-1) Q/B (Year-1) EE P/Q WHB Polar Bear 4.857 0.0005 0.129 2.080 0.414 0.062 SH Polar Bear 4.906 0.0004 0.154 2.080 0.506 0.074 Polar Bear Foxe 4.927 0.0002 0.121 2.080 0.304 0.058 Killer Whale 4.872 0.0000 0.151 4.998 0.265 0.030 Narwhal 4.062 0.0019 0.084 26.182 0.271 0.003 Bowhead 3.335 0.0109 0.021 5.475 0.384 0.004 Walrus N 3.332 0.0027 0.172 47.123 0.188 0.004 Walrus S 3.452 0.0010 0.097 33.778 0.143 0.003 Bearded Seal 3.866 0.0037 0.176 14.262 0.791 0.012 Harbour Seal 3.971 0.0010 0.125 18.612 0.074 0.007 Ringed Seal 4.077 0.0469 0.158 17.272 0.413 0.009 Harp seal 4.103 0.0010 0.126 15.660 0.688 0.008 Beluga E 3.694 0.0021 0.066 21.448 0.220 0.003 Beluga W 3.873 0.0247 0.064 16.713 0.133 0.004 Beluga James 3.869 0.0015 0.087 16.623 0.679 0.005 Seabirds 3.839 0.0650 0.370 17.258 0.950 0.021 Arctic Char 3.300 0.412 0.200 1.500 0.950 0.133 Atlantic Salmon 3.450 0.148 0.520 7.150 0.950 0.073 Gadiformes 3.235 0.853 0.470 1.850 0.950 0.254 Sculpins/ Zoarcids 3.188 0.382 0.700 3.269 0.950 0.214 Capelin 3.132 0.488 1.700 4.800 0.950 0.354 Sandlance 3.128 0.705 0.850 3.450 0.950 0.246 Sharks/Rays 4.033 3.18E-06 0.220 1.250 0.950 0.176 Other Marine Fish 2.948 0.374 0.932 3.018 0.950 0.309 Brackish Fish 3.216 0.055 3.500 5.798 0.950 0.604 Cephalopods 3.645 0.227 1.500 5.000 0.950 0.300 MacroZooplankton 2.711 7.5000 1.000 3.000 0.278 0.333  Euphausiids 2.787 2.1480 3.300 13.200 0.800 0.250 Copepods 2.050 4.0150 16.000 64.000 0.472 0.250 Crustaceans 2.410 1.8000 3.600 14.400 0.584 0.250 Other MesoZooplankton 2.336 1.2100 10.000 40.000 0.556 0.250 MicroZooplankton 2.000 2.235 15.000 45.000 0.950 0.333 Marine Worms 2.275 5.930 0.600 4.000 0.950 0.150 Echinoderms 2.575 8.708 0.300 1.000 0.950 0.300 Bivalves 2.148 5.942 0.570 6.300 0.950 0.091 Other Benthos 2.091 3.139 2.500 12.500 0.950 0.200 Pelagic Production 1.000 8.0000 46.865 0.000 0.800 ‐ Ice Algae 1.000 3.5000 46.197 0.000 0.650 ‐ Ice Detritus 1.000 0.0090 ‐  ‐  0.904 ‐ Detritus 1.000 0.3300 ‐  ‐  0.224 ‐      From the Tropics to the Poles, Wabnitz and Hoover  Monte Carlo Results Monte Carlo simulations were run using the pedigree ranking from Ecopath version 5 (Christensen et al. 2005). C.V. values were estimated based on quality of input data (see appendix 6 for all CV values and appendix 7 for graphs of biomass and P/B results). MC simulations were unable to improve the sum of squares value obtained by fitting the model. However, ranges of plausible ranges were obtained for biomass and P/B parameters.  Biomass input CV and output with limits are presented in table 8. Most marine mammal biomass results remained quite close to the starting value. Ringed seals had the largest starting biomass of any marine mammal group, and also the highest upper limit or largest biomass which could be supported by the system, followed by WHB Bay beluga and bowhead whales. Ringed seals had a large uncertainty, as population sizes are not well known, however the model is able to support a large biomass of these seals. Within the model framework, bowheads have the potential to double the biomass and still be supported by the ecosystem.  Although there was high uncertainty with the biomass of fish groups, the ability of the system to sustain moderate biomasses of fish is an added discovery due to the understudied nature of fish within the ecosystem. While commercial fishing endeavours have not been profitable, it would be assumed the region has a conservative fish biomass. Compared to other ecosystem models, total fish biomass is lower than other systems of similar latitude. Total fish biomass of HB is 3.42 t·km-2 compared to 4.32 t·km-2 in the Antarctic Peninsula (Hoover unpublished data), although the Antarctic is more productive, the dominant species is krill (Euphausia superba), and commercial fisheries operations in this region have also proved difficult. Table 8: CV used for Monte Carlo estimates of biomass. Results show the mean biomass, along with the upper and lower limits of the 95% CI presented in t·km-2    Functional Group Biomass (CV) Lower Limit Mean Biomass Upper Limit 1 Polar Bear WHB 0.15 0 0 0.001 2 SH Polar Bear 0.15 0 0 0 3 Polar Bear Foxe 0.15 0 0 0 4 Killer Whale 0.15 0 0 0 5 Narwhal 0.15 0.001 0.002 0.003 6 Bowhead 0.4 0.002 0.011 0.02 7 Walrus N 0.25 0.001 0.003 0.004 8 Walrus S 0.25 0 0.001 0.001 9 Bearded Seal 0.25 0.002 0.004 0.006 10 Harbour Seal 0.25 0.001 0.001 0.002 11 Ringed Seal 0.25 0.023 0.047 0.07 12 Harp seal 0.25 0.001 0.001 0.002 13 Beluga E 0.15 0.001 0.002 0.003 14 Beluga W 0.15 0.017 0.025 0.032 15 Beluga James 0.15 0.001 0.001 0.002 16 Seabirds 0.4 0.013 0.065 0.117 17 Arctic Char 0.1 0.329 0.412 0.494 18 Atlantic Salmon 0.1 0.118 0.148 0.177 19 Gadiformes 0.1 0.683 0.853 1.024 20 Sculpins/ Zoarcids 0.1 0.305 0.382 0.458 21 Capelin 0.1 0.39 0.488 0.585 22 Sandlance 0.1 0.564 0.705 0.846 23 Sharks/Rays 0.1 0 0 0 24 Other Marine Fish 0.1 0.3 0.374 0.449 25 Brackish Fish 0.1 0.044 0.055 0.066 26 Cephalopods 0.25 0.113 0.227 0.34 27 Macro-Zooplankton 0.25 3.75 7.5 11.25 28 Euphausiids 0.15 1.504 2.148 2.792 29 Copepods 0.15 2.811 4.015 5.22 30 Crustaceans 0.15 1.26 1.8 2.34 31 Other Meso-Zoopl. 0.15 0.847 1.21 1.573 32 Micro-Zooplankton 0.25 1.117 2.235 3.352 33 Marine Worms 0.1 4.744 5.93 7.115 34 Echinoderms 0.1 6.966 8.708 10.449 35 Bivalves 0.1 4.753 5.942 7.13 36 Other Benthos 0.1 2.511 3.139 3.767 37 Primary Production 0.15 5.6 8 10.4 38 Ice Algae 0.15 2.45 3.5 4.55 36  An Ecosystem Model of Hudson Bay, Canada    Total zooplankton biomass of 18.91 t·km-2 appears to fall within the ranges of observed samples. Harvey et al. (2006), estimated macro and meso-zooplankton from >1 to 6 g DW m-2 for central HB, while a few samples from Harvey et al. (2001) reached close to 10 g DW m-2 in northern HB. Using the conversion of DW:WW of 5 (Cushing et al. 1958), this would indicate values of zooplankton biomass from 5-30 g·m-2 within Hudson Bay and up to 50 g·m-2  for Hudson Strait. However these high values were obtained from late summer values, and are likely not representative of an annual value. Fitting Results Results of time series fitting, including effort and mediation are presented in figure 10. While most trends were captured by the model, there were a few exceptions. Foxe Basin polar Bear catch was not forced due to the unknown portion of catches coming from within the model area. Therefore it was presented as a relative catch sequence. Although the values for the data and the model are not the same, the trend appears to be similar, with catches decreasing and levelling out by the late 1980s. James Bay beluga abundance was not able to increase to levels as high as survey estimates. While migration from the EHB beluga group (through biomass accumulation) improved the fit for both EHB and James Bay belugas, the full magnitude of the increase was unable to be fully captured within the model. Data for fitting fish groups provided insight as to general trends of abundance; however the model was unable to simulate the extreme increase in capelin and sandlance populations, as well as the full decreases in gadiformes and  sculpins/zoarcids. Figure 10: Time series data (open dots) with abundance and catch trends (lines) from model output. From the Tropics to the Poles, Wabnitz and Hoover  Biomass accumulation was crucial to obtaining fits for bowhead and EHB belugas. Bowheads were unable to increase as rapidly within the model, starting at such a low biomass, and a low P/B. Conversely, a small decline in EHB belugas was created through hunting mortality and vulnerability settings, but was not fully captured until a negative biomass accumulation component was added.  All polar bear groups demonstrated stable population sizes with hunting pressure. Vulnerabilities were able to cause small increases or decreases in the populations, however, the addition of mediation increased the sensitivity of these groups to changes in sea ice as well as vulnerabilities of their prey. Once the mediation function was applied (to arena area and vulnerability of prey), all polar bear groups became highly sensitive to small changes in vulnerabilities.  Starting from the bottom of the food web, shifts caused by forcing functions can be identified. Figure 11 identifies changes in the lowest trophic levels of the ecosystem, with declines in ice algae and ice detritus of nearly 10% each, and increases in pelagic production (26%), and pelagic detritus (33%). Since both the ice algae and the pelagic production groups were forced, these changes are not surprising.  Changes in the detritus and producers are propagated to the next trophic levels, as shown in figure 12 by declines in all benthic groups, with the exception of crustaceans (although this group contains pelagic and benthic crustaceans). Zooplankton, however, fare much better, with increases ranging from 12% (micro-zooplankton) to 58% (macro-zooplankton). The increase in zooplankton is caused by the diets containing large concentrations of pelagic production, which supersede the declines in the ice algae contribution of the diet.  Declines are identified predominantly in benthic fish (Gadiformes: Arctic and Polar cod, Sculpins/Zoarcids: benthic fish, and sharks/rays) due to diets consisting of ice detritus and other benthos (figure 13). Gadiformes and sculpins/zoarcids decreased in the diet of thick-billed murres an average of 68 and 57%, respectively (Gaston et al. 2003)6. Pelagic based fish show increases, with the largest being capelin and sandlance. Fitting of time-series data (figure 10) from the diet of thick-billed murres, appears to be unable to capture the full magnitude of the increase for both capelin and sandlance. Capelin increased in the diet from 20 to 50%, and sandlance from 4 to 20% (as averaged from the first and last 3 years). In the model these groups show substantial increases with capelin increasing over 70% of their original biomass, with                                                  6 When comparing the average contribution to the diet of thick billed murres as the average value for the first and last 3 years of the diet study. Figure 11: Model end biomass for 2010 presented as percentage change from starting biomass for producers and detritus. Figure 12: Model ending biomass for 2010 presented as percentage change from the starting biomass for zooplankton and benthic groups 38  An Ecosystem Model of Hudson Bay, Canada    sandlance nearly doubling. Increased hunting and fishing pressure for birds and fish groups does not appear to be causing declines, as the mortality caused by hunting and fishing for these groups was quite small in relation to total mortality (table 2). Seabird biomass was still able to increase within the model, despite hunting effort increasing roughly 4.5 times the 1970 effort.  Most marine mammal functional groups were fit to abundance data, therefore changes in biomass were previously known. All polar bear groups declined in biomass (figure 14); primarily due to the mediation function hindering their ability to hunt effectively when there is less sea ice. Narwhal decreases are due to increasing hunting mortality. Biomass for narwhal remains relatively stable from 1970-2000. However, when catches are increased from 2000-2010, the population begins to decline, but only in the last 10 years of the simulation, indicating this is the result of hunting pressure (see narwhal graph in appendix 8). Removal of catches in the model identifies an increase in narwhal biomass Bearded seals also appear to decline due to hunting mortality (figure 14). For bearded seals, hunting mortality accounts for one third of all mortality in Ecopath. Combined with the increases in human population and hunting pressure, by 2010 the hunting mortality is nearly 10 times the predation mortality indicating harvest of bearded seals is causing the decline within the model. The harp seal group also shows hunting mortality to increase to double the predation mortality by the end of the simulation. However, because catches for this group were set low in the first year, large increases in catch are still unable to cause a decline overall. Ringed and harbour seals show low hunting mortality throughout the simulation, indicating the populations are large enough to sustain the effort levels used in the model fitting.  Both walrus groups (N and S) experience less predation from polar bears, due to declining populations. N walrus increase in the model due to low harvest levels from 2003-2010. S walrus experienced higher hunting pressure during this time, causing the decrease observed at the end of the model simulation.  Killer whale abundance was forced within the model to replicate the observed increase in killer whales. Biomass accumulation was unable to explain this large increase, leading the authors to believe the changes may be caused by immigration from other areas. Figure 13: Model ending biomass for 2010 presented as percentage change from starting biomass for fish and seabirds. Figure 14: Model ending biomass for 2010 presented as percentage change from starting biomass for marine mammals.  From the Tropics to the Poles, Wabnitz and Hoover  DISCUSSION There are many dynamics expected to alter Arctic ecosystems presently and in the future; climate change, environmental contaminants, off-shore oil and gas activities, shipping, hunting, and commercial fisheries (Huntington 2009). While there are no models large enough to capture the dynamics of Arctic ecosystems including all of these potential threats, this model brings us closer through the inclusion of climate dynamics, and hunting. There are most certainly more dynamics shaping the Hudson Bay ecosystem than included in this model, however there is much to be learned before we move on to incorporating more influences.  Parkinson et al. (1999) found a slight (although not significant) negative trend in sea ice from 1979-1996, along with positive ice thickening trends in Hudson Bay. Although the timing of the spring melt has moved earlier in the year -0.49-1.25 days year-1, and freeze-up has shifted later 0.32-0.55 days year-1 (Gagnon and Gough 2005). Gough et al. (2004) showed a decrease in Arctic summer ice extent has decreased 15-30% in the last 30 years. This model captures a 10% decrease in ice algae biomass, assuming a linear relationship between ice algae and extent of sea ice, however this decline is not as large as the Arctic summer average decline of 15-30%. As ice declines in HB have shown to be less intense than other Arctic areas, the declines used to force the model (and the ice algae group) appear to be within range of observed changes; less than the Arctic average, but still showing declines. If sea ice declines escalate to be on par with high Arctic areas, it is likely biomass changes observed in the model would be exaggerated.  Bentho-pelagic coupling of sea ice to ice detritus may be an important factor in determining the abundance of benthic communities. Damaged algal cells from the sea ice sink faster than healthy ones, and increased runoff flushing algal cells through brine channels in the ice also increase exports (Tremblay et al. 1989). Export of ice algae to the benthic community was estimated at a minimum of 20% in HB (Tremblay et al. 1989). Moreover, accumulation of algal biomass within the sea ice is thought to favor an effective transfer to the benthos, as aggregated algal cells sink up to three times faster than individual algal cells (Riebesell et al. 1991). This would indicate changes in the HB sea ice could have a strong impact on the benthic community. It should be noted that the timing of ice melt generally coincides with the pelagic bloom, making for a complex dynamic in bentho-pelagic coupling (Smith et al. 2006). Decreases in benthic groups were observed in the model. These decreases in benthos were impacted by declining ice, yet there are certainly other factors in the natural environment. In the model, these changes further explain the decreases in benthic fish (as reported from thick-billed murre diets). If the bentho-pelagic coupling was disrupted, it would allow for restructuring of the ecosystem where pelagic species would dominate lower trophic levels. In fact we see increases in zooplankton biomasses, as they benefit from increased spring blooms.  Zooplankton may not continue to thrive under increasingly warming conditions. As temperature increases, so will river runoff and freshwater inputs to the system, causing both increased nutrients and increased stratification in the water column. However, the impacts to the zooplankton community as a whole remain unknown.  As thick-billed murre diets indicated, there was a shift from Arctic to sub-Arctic fish composition; from cod, sculpins, and zoarcids to capelin and sandlance (Gaston et al. 2003). Although it appears that trends were not fully captured (figure 10) large changes in biomass (figure 13) indicate substantial changes in biomass have occurred. As there is a gross lack of data on fish populations in this region, diets of birds was the only indication of fish abundance. Because the data are based on the northern edge of the region, it is questionable as to the capacity to be extrapolated to the entire region. Fish in southern HB are likely to be impacted differently with large freshwater inputs from rivers, causing different environmental conditions. These possibilities need to be explored along with basic comprehensive surveys in order to provide more accurate modelling of fish groups.  Hunting and fishing data is important in understanding the human impacts upon species. Significant under-reporting is likely to have occurred throughout the region, especially in the unregulated catches of seals, birds, and fish. While in the model the mortality inflicted on these groups is nearly negligible, in reality the subsistence harvest is almost certainly an important contribution to the mortality of most groups. The exception will most likely be ringed seals, as the biomass in the region is large and less likely 40  An Ecosystem Model of Hudson Bay, Canada    to be impacted by the number of human inhabitants. Ultimately, better estimates of subsistence hunting would lead to better overall indication of the pressures upon marine mammals, fish, and birds.  Marine mammal results provide insight as to potential reasons for changes in abundance, because trends are already known. James Bay beluga increases were not fully captured within the model, even with migration from the EHB beluga population. If the JB population is a mixture of eastern and western HB beluga population, migration from WHB belugas could potentially contribute enough migration to simulate the increase. Another possibility is the reported increase is a factor of survey methods (or survey coverage) rather than actual increases. However, this will likely never be revealed, so at present the increase is believed to be occurring, yet not fully explained by the model.  The narwhal of HB, once believed to be part of a larger population, are under growing hunting pressure, with catches increasing in recent years. As identified in the model, larger harvests are already causing declines in the population, with catches increasing since 2000.  Declines in polar bear populations were difficult to simulate through trophic interactions and hunting harvests. However, the inclusion of a mediating function based on sea ice was key to capturing these declines. This information was necessary to improve the model, however because it is well documented (Stirling and Derocher 1993; Stirling et al. 1999; Stirling and Parkinson 2006) we were able to incorporate the indirect effect of sea ice. What remains unknown are the other potentially important interactions between species and their habitat which have yet to be studied, most notable the lower trophic levels.  Future research on the HB ecosystem should focus on expanding our understanding of lower trophic level organisms such as zooplankton and fish. While zooplankton have been sampled in the past, HB is lacking any formal surveys of fish populations. In addition, increased samples will start to identify temporal and spatial trends of these species. Diets of lower level organisms are poorly understood, and the basis of the food web is contingent upon this understanding. More research is needed to better understand the impacts changes in the food web have on higher trophic level species (marine mammals), and to improve the model. If in fact changes to fish populations are occurring at the same rate in the ecosystem as they are in the diets of birds, then the issue of bentho-pelagic coupling should be explored with fish and zooplankton in mind. There will ultimately be consequences to top predators, however if we wait to identify these changes at higher trophic levels, it is almost certain large scale shift will have occurred throughout the system. CONCLUSIONS The Hudson Bay ecosystem model summarizes the information known about the species which inhabit the region. Through the exercise of creating a model and fitting it to data, important linkages in the system are identified or validated if previously known. A well-documented linkage further validated by the model is that sea ice is an essential factor in determining the abundance of polar bears. Bentho-pelagic coupling of sea ice to benthos via the spring melt is identified as a potentially vital link in determining the future of Hudson Bay as a benthic or pelagic dominated ecosystem. However, our general lack of knowledge for lower trophic level species (benthos, zooplankton, and fish) may be the greatest hindrance in terms of modelling and expanding our knowledge on this ecosystem.   From the Tropics to the Poles, Wabnitz and Hoover  ACKNOWLEDGEMENTS Funding for this project was provided by the Global Warming and Arctic Marine Mammal (GWAMM) project under the International Polar Year (IPY) funding. Without Steve Ferguson, the principal investigator for the GWAMM project, this research would have never presented itself. A special thanks to members of this project for their collaboration and input of data regarding Hudson Bay; Jeff Higdon, Elly Chmelnitsky, Magaly Chambellant, and Sebastian Luque. To the numerous researchers who assisted me with expert knowledge when information was needed; Margaret Tresle, Jim Reist, Ross Tallman, Pierre Richard, Lisa Loseto, Tara Bortoluzzi, David Barber and the researchers at CEOS in Winnipeg, and especially Bruce Stewart. Thanks to Megan Bailey, Brooke Campbell, Jeroen Steenbeek, Divya Varkey, Chiara Piroddi, Laura Tremblay-Boyer, and Colette Wabnitz for technical assistance. Finally, travel to Hudson Bay was provided by the Cecil and Kathleen Morrow Scholarship, and provided unparalleled insight into the dynamic of the ecosystem through first-hand experience, thank you.      42  An Ecosystem Model of Hudson Bay, Canada    REFERENCES Aars, J., Lunn, N. J. and Derocher, A. E. (2005). Polar Bears: Proceedings of the 14th working meeting of the IUCN/SSC Polar bear specialist groups. ICUN: The World Conservation Union. Seattle, WA: v+188. American Cetacean Society (2004). "American Cetacean Society Fact Sheet: Bowhead whale (Balaena mysticetus)." Retrieved March 23, 2010, from http://www.acsonline.org/factpack/bowhead.htm.  Angliss, R. P. and Outlaw, R. B. (2006). The bearded seal (Erignathus barbatus): Alaska stock. Alaska Marine Mammal Stock Assessments 2006. NOAA: 1-4.  Anonymous (1979). Research to establish present levels of marine harvesting: Harvest by the Inuit of northern Quebec- phase II (Year 1976). Montreal, James Bay and Northern Quebec Native Harvesting Research Committee: xv+99.  Arctic Climate Impact Assessment (2004). Impacts of a warming Arctic. C. U. Press. New York: 139.  Atkinson, A., Meyer, B., Stübing, D., Hagen, W., Schmidt, K. and Bathman, U. V. (2002). "Feeding and energy budgets of Antarctic krill Euphausia superba at the onset of winter-II. Juvenile and adults." Limnology and oceanography 47(4): 953-966. Atkinsor, E. G. and Wacasey, J. W. (1989). Benthic invertebrates collected from Hudson Bay, Canada, 1953 to 1965. Canadian Data Report of Fisheries and Aquatic Science. Department of Fisheries and Oceans. Ste Anne de Bellevue, QB: iv+ 121. Barlow, J. and Boveng, P. (1991). "Modeling age-specific mortality for marine mammal populations." Marine Mammal Science 7(1): 50-65. Bathmann, U. V., Makarov, R. R. and Rohardt, G. (1993). "Winter distribution and overwintering strategies of the Antarctic copepod species Calanoides acutus, Rhincalanus gigas and Calanus propinquus (Crustacea, Calanoidea) in the Weddell Sea." Polar Biology 13(5): 333-346. Beck, G. G., Hammil, M. O. and Smith, T. G. (1993). "Seasonal variation in the diet of harp seals (Phoca groenlandica) from the Gulf of St Lawrence and western Hudson Strait." Canadian Journal of Fisheries and Aquatic Sciences 50: 1363-1371. Bell, J. (2002). Nunavut's high population growth rate slows. Nunatsiaq Online. Retieved March 15, 2010 from http://www.nunatsiaqonline.ca/archives/nunavut020315/news/nunavut/20315_2.html  Bigg, M. A. (1981). Harbour seal- Phoca vitulina and P. larga. Handbook of marine mammals. S. H. Ridgway and R. J. Harrison. London, Academic Press. 2: 1-27. Booth, S. and Watts, P. (2007). Canada's arctic marine fish catches. Reconstruction of marine fisheries catches for key countries and regions (1950-2005). D. Zeller and D. Pauly, Fisheries Centre Research Reports. 15(2): 3-15. Born, E. W., Rysgaard, S., Ehlmé, G., Sejr, M., Acquarone, M. and Levermann, N. (2003). "Underwater observations of foraging free-living Atlantic walruses (Odobenus rosmarus rosmarus) and estimates of their food consumption." Polar Biology 26: 348-357. Brey, T. and Gerdes, D. (1998). "High Antarctic macrobenthic community production." Journal of Experimental Marine Biology and Ecology 231: 191-200. British Atmospheric Data Centre (2010). HasISST 1.1 Global Sea Ice coverage and SST 1980-Present, British Atmospheric Data Centre, 1.1. Brock, M. J. and Miller, D. C. (1999). "Particle selectivity, gut volume, and the response to a step change in diet for deposit-feeding polychaetes." Limnology and Oceanography 44(4): 1132-1138. Burns, J. J. (2002). Harbor Seal and Spotted Seal (Phoca vitulina and P. largha). The Encyclopedia of Marine Mammals. W. F. Perrin, B. Würsig and J. G. M. Thewissen. San Diego, Academic Press: 552-560. Buszowski, J., Christensen, V., Gao, F., Hui, J., Lai, S., Steenbeek, J., Walters, C. and Waters, W. (2007). Ecopath with Ecosim 6, University of British Columbia. Calbert, A. (2001). "Mesozooplankton grazing effect on primary production: A global comparative analysis in marine ecosystems." Limnology and Oceanography 46(7): 1824-1830.  Cauffopé, G. and Heymans, S. J. J. (2005). Energy contents and conversion factors for sea lion's prey. Food web models and data for studying fisheries and environmental impacts on Eastern Pacific ecosystems. S. Guénette and V. Christensen, Fisheries Centre Research Reports. 13(1): 225-237. Chambellant, M. (2010). Hudson Bay ringed seal: Ecology in a warming climate. A Little Less Arctic: Top Predators in the World's Largest Northern Inland Sea, Hudson Bay. S. Ferguson, L. Loseto and M. Mallory, Springer: 137-158. Chambellant, M. and Ferguson, S. H. (2009). Comparison of strip- and line-transect sampling to estimate density and abundance of ringed seals (Phoca hispida) in western Hudson Bay. Canadian Science Advisory Secretariat. Department of Fisheries and Oceans. Winnipeg, Canada: 1-23. Christensen, V., Walters, C., Pauley, D. and Forrest, R. (2007). Ecopath with Ecosim version 6.0. User manual/ help files guide. Lenfest Oceans Futures Project 2007. University of British Columbia, British Columbia, Canada. version 2.0. Christensen, V., Walters, C. and Pauly, D. (2005). Ecopath with Ecosim: a user's guide. Vancouver, BC, Fisheries Centre: University of British Columbia: 1-154. From the Tropics to the Poles, Wabnitz and Hoover  Cosens, S. E. and Innes, S. (2000). "Distribution and numbers of Bowhead whales (Balaena mysticetus) in Northwestern Hudson Bay in August 1995." Arctic 53(1): 36-41. COSEWIC (2004a). COSEWIC assessment and update status report on the beluga whale Delphinapterus leucas in Canada. Committee on the Status of the Endangeres Wildlife in Canada. Ottawa, Canada: ix+70. COSEWIC (2004b). COSEWIC assessment and update status report on the narwhal Monodon monoceros in Canada. Committee on the Status of the Endangeres Wildlife in Canada. Ottawa: vii+50. COSEWIC (2006). Assessment and update status report on the Atlantic walrus Odobenus rosmarus rosmarus in Canada Committee on the Status of the Endangeres Wildlife in Canada. Ottawa, Canada: ix+65. Cripps, G. C. and Atkinson, A. (2000). "Fatty acid composition as an indicator of carnivory in Antarctic krill, Euphausia superba." Canadian Journal of Fisheries and Aquatic Sciences 57(Supplement 3): 31-37. Culik, B. M. (2004). Review of Small Cetaceans: Distribution, Behaviour, Migration and Threats. Marine Mammal Action Plan. Convention on Migratory Species. Bonn, Germany, United Nations Environment Programme: 343. Cushing, D. H., Humphrey, G. F., Banse, K. and Leaevastu, T. (1958). "Report of the committee on terms and equivalents." Rapports et Procès-verbaux des Réunions du Conseil Permanent International pour l'Exploration de la Mer 14(4): 15-16. Danielson, E. W. (1971). "Hudson Bay ice conditions." Arctic 24(2): 90-107. Dauby, P., Scailteur, Y., Chapelle, G. and De Broyer, C. (2001). "Potential impact of the main benthic amphipods on the eastern Weddell Sea shelf ecosystem (Antarctica)." Polar Biology 24: 657-662. De March, B. G. E. and Postma, L. D. (2003). "Molecular genetic stock discrimination of belugas (Delphinapterus leucas) hunted in eastern Hudson Bay, Northern Quebec, Hudson Strait, and Sanikiluaq (Belcher Islands), Canada, and comparisons to adjacent populations." Arctic 56(2): 111-124. Dempson, B. J., Shears, M. and Bloom, M. (2002). "Spatial and temporal variability in the diet of anadromous Arctic carr, salvelinus alpinus, in northern Labrador." Environmental Biology of Fishes 64: 49-62. Derocher, A. E., Andriashek, D. and Stirling, I. (1993). "Terrestrial foraging by polar bears during the ice-free period in western Hudson Bay." Arctic 46(3): 251-254. DFO (1990). Annual summary of fish and marine mammal harvest data for the Northwest Territories 1988-1989, Volume 1. Department of Fisheries and Oceans. Winnipeg, MB, Freshwater Institute, Central and Arctic Region: 59. DFO (1991). Annual summary of fish and marine mammal harvest data for the Northwest Territories 1989-1990, Volume 2. Department of Fisheries and Oceans. Winnipeg, MB, Freshwater Institute, Central and Arctic Region: 61. DFO (1992). Annual summary of fish and marine mammal harvest data for the Northwest Territories 1990-1991, Volume 3. Department of Fisheries and Oceans. Winnipeg, MB, Freshwater Institute, Central and Arctic Region: 67. DFO (1993). Annual summary of fish and marine mammal harvest data for the Northwest Territories 1991-1992, Volume 4. Department of Fisheries and Oceans. Winnipeg, MB, Freshwater Institute, Central and Arctic Region: 69. DFO(1994). Annual summary of fish and marine mammal harvest data for the Northwest Territories 1992-1993, Volume 5. Department of Fisheries and Oceans. Winnipeg, MB, Freshwater Institute, Central and Arctic Region: 104. DFO (1995). Annual summary of fish and marine mammal harvest data for the Northwest Territories 1993-1994, Volume 6. Department of Fisheries and Oceans. Winnipeg, MB, Freshwater Institute, Central and Arctic Region: 86. DFO (1996). Annual summary of fish and marine mammal harvest data for the Northwest Territories 1994-1995, Volume 7. Department of Fisheries and Oceans. Winnipeg, MB, Freshwater Institute, Central and Arctic Region: 85. DFO (1997). Annual summary of fish and marine mammal harvest data for the Northwest Territories 1995-1996, Volume 8. Department of Fisheries and Oceans. Winnipeg, MB, Freshwater Institute, Central and Arctic Region: 80. DFO (1998). Hudson Bay narwhal. Stock status report E5-44. Department of Fisheries and Oceans, Central and Arctic Region: 4. DFO (1999). Hudson Bay-Foxe Basin bowhead whales. Science. Department of Fisheries and Oceans, Department of Fisheries and Oceans Central and Arctic: 8. DFO (2001). Northern Quebec (Nunavik) beluga (Delphinapterus leucas). Stock Status Report E4-01 (2001). Department of Fisheries and Oceans: 1-7. DFO (2002a). Atlantic walrus. Department of Fisheries and Oceans. Winnipeg, Canada: 1-19. DFO (2002b). The Beluga. Communications Directorate. Department of Fisheries and Oceans. Ottawa, Canada. DFO (2008). "Aquatic Species at Risk - Beluga Whale (Eastern Hudson Bay)." Retrieved May 8, 2010, from http://www.dfo-mpo.gc.ca/species-especes/species-especes/beluga-Hudson-eng.htm. DFO (2009). Review of aerial survey estimates for ringed seals (Phoca hispida) in western Hudson Bay. Canadian Science Advisory Secretariat. Department of Fisheries and Oceans, Central and Arctic Region: 5. Dunbar, M. J. (1983). "A unique International Polar Year contribution: Lucien Turner, capelin, and climate change." Arctic 36(2): 204-205. 44  An Ecosystem Model of Hudson Bay, Canada    Efran, A. and Pitcher, T. J. (2005). An ecosystem model simulation of the Antarctic Peninsula. Modeling Antarctic Marine Ecosystems. M. L. D. Palomares, P. Pruvost, T. J. Pitcher and D. Pauley. University of British Columbia, Vancouver, Canada, Fisheries Centre Research Reports. 13(7): 5-20. Fabijian, M. and Usher, P. J. (2003). Inuvialuit harvest study data and methods report 1988-1997, The Inuvialuit Joint Secretariat: v + 202  Farabee, M. J. (2001). "Chlorophyll a." Retrieved 03/08/10, 2010, from http://www.emc.maricopa.edu/faculty/farabee/BIOBK/BioBookPS.html. Ferguson, S. H. (2007). Assessment of eastern Arctic bowhead whales (Balaena mysticetus). Canadian Science Advisory Secretariat. Department of Fisheries and Oceans. Winnipeg, Canada: 1-10. Finley, K. J. and Evans, C. R. (1983). "Summer diet of the bearded seal (Erignathus barbatus) in the Canadian Arctic." Arctic 36(1): 82-89. Fisher, K. I. and Stewart, R. E. A. (1997). "Summer foods of Atlantic walrus, Odobenus rosmarus rosmarus, in northern Foxe Basin, Northwest Territories." Canadian Journal of Zoology 75: 1166-1175. Ford, J. K. B. (2002). Killer Whale (Orcinus orca). The Encyclopedia of Marine Mammals. W. F. Perrin, B. Würsig and J. G. M. Thewissen. San Diego, Academic Press: 669-676. Froese, R. and Pauly, D. (2008). "FishBase." World Wide Web electronic publication. www.fishbase.org, version (12/2008). Gagnon, A. S. and Gough, W. A. (2005). "Trends in the dates of ice freeze-up and breakup over Hudson Bay, Canada." Arctic 58(4): 370-382. Gamble, R. L. (1988). Native harvest of wildlife in the Keewatin Region, Northwest Territories for the period October 1985 to March 1986 and a summary for the entire period of the harvest study from October 1981 to March 1986. Canadian Data Report of Fisheries and Aquatic Sciences. 688: v+85. Garde, E., Heide-Jørgensen, M. P., Hansen, S. H., Nachman, G. and Forchhammer, M. C. (2007). "Age-specific growth and remarkable longevity in narwhals (Monodon monoceros) from west Greenland as estimated by aspartic acid racemization." Journal of Mammalogy 88(1): 49-58. Garrison, D. L. and Buck, K. R. (1989). "The Biota of Antarctic pack ice in the Weddell Sea and the Antarctic Peninsula regions." Polar Biology 10: 211-219. Gaston, A. J., Bertram, D. F., Boyne, A. W., Chardine, J. W., Davoren, G., Diamond, A. W., Hedd, A., Montevecchi, W. A., Hipfner, J. M., Lemon, M. J. F., Mallory, M. L., Rail, J.-F. and Robertson, G. J. (2009a). "Changes in Canadian seabird populations and ecology since 1970 in relation to changes in oceanography and food webs." Environmental Reviews 17: 267-286. Gaston, A. J., Gilchrist, H. G., Mallory, M. L. and Smith, P. A. (2009b). "Changes in seasonal events, peak food availability, and consequent breeding adjustment in a marine bird: a case of progressive mismatch." The Condor 111(1): 111-119. Gaston, A. J., Woo, K. and Hipfner, J. M. (2003). "Trends in forage fish populations in northern Hudson Bay since 1981, as determined from the diet of nestling thick-billed murres Uria lomvia." Arctic 56(3): 227-233. George, J. C., Bada, J., Zeh, J., Scott, L., Brown, S. E., O'hara, T. and Suydam, R. (1999). "Age and growth estimates of bowhead whales (Balaena mysticetus) via aspartic acid racemization." Canadian Journal of Zoology 77: 571-580. Gerdiner, K. and Dick, T. A. (2010). "Arctic cephalopod distribution and their associated predators." Polar Research 29: 209-227. Gilchrist, H. G. and Robertson, G. J. (1999). "Population trends of gulls and Arctic terns nesting in the Belcher Islands, Nunavut." Arctic 52(4): 325-331. Gonzalez, N. (2001). Inuit traditional ecological knowledge of the Hudson Bay narwhal (Tuugaalik) population. Department of Fisheries and Oceans. Iqaluit, Nunvut, Canada. Gosselin, J.-F. (2005). Abundance indicies of belugas in James Bay and eastern Hudson Bay in summer 2004. Canadian Science Advisory Secretariat. Department of Fisheries and Oceans. Mont-Joli, Canada: i+22. Gosselin, J.-F., Lesage, V. and Hammil, M. O. (2009). Abundance indicies of beluga in James Bay, eastern Hudson Bay and Ungava Bay in 2008. Canadian Science Advisory Secretariat. Department of Fisheries and Oceans Canada. Mont-Joli, Canada: iv+25. Gosselin, J.-F., Lesage, V., Hammil, M. O. and Bourdages, H. (2002). Abundance indices of beluga in James Bay, eastern Hudson Bay and Ungava in the summer 2001. Canadian Science Advisory Secretariat. Department of Fisheries and Oceans: 1-28. Gosselin, M., Levasseur, M., Wheeler, P. A., Horner, R. A. and Booth, B. C. (1997). "New measurements of phytoplankton and ice algal production in the Arctic Ocean." Deep Sea Research II 44(8): 1623-1644. Gough, W. A., Cornwell, A. R. and Tsuji, L. J. S. (2004). "Trends in seasonal sea ice duration in southwestern Hudson Bay." Arctic 57(3): 299-305. Guénette, S. (2005). Models of southeast Alaska. Food web models and data for studying fisheries and environmental impacts on Eastern Pacific ecosystems. S. Guénette and V. Christensen, Fisheries Centre Research Reports. 13(1): 106-179. Hammill, M. O. (2001). Beluga in northern Quebec: Impact of harvesting on population trends in eastern Hudson Bay. Canadian Science Advisory Secretariat research document; 2001/025, Department of Fisheries and Oceans: 18. From the Tropics to the Poles, Wabnitz and Hoover  Hammill, M. O., Kingsley, M. C. S., Lesage, V. and Gosselin, J.-F. (2009). Abundance of eastern Hudson Bay belugas. Canadian Science Advisory Secretariat. Department of Fisheries and Oceans Canada. Mont-Joli, Canada: iv+22. Hansen, J., Sato, M., Ruedy, R., Lo, K., Lea, D. W. and Medina-Elizade, M. (2006). "Global temperature change." Proceedings from the National Academy of Sciences of the United states of America 103(39): 14288-14293. Harvey, M., Starr, M., Therriault, J. C., Saucier, F. and Gosselin, M. (2006). "MERICA-Nord Program: Monitoring and researching the Hudson Bay complex." Atlantic Zone Monitoring Program Bulletin 5: 27-32. Harvey, M., Therriault, J.-C. and Simard, N. (2001). "Hydrodynamic control of late summer species composition and abundance of zooplankton in Hudson Bay and Hudson Strait (Canada)." Journal of Plankton Research 23(5): 481-496. Harvey, M., Therriault, J. C. and Simard, N. (1997). "Late-summer distribution of phytoplankton in relation to water mass characteristics in Hudson Bay and Hudson Strait (Canada)." Canadian Journal of Fisheries and Aquatic Sciences 54: 1937-1952. Harwood, L. A., Norton, P., Day, B. and Hall, P. A. (2002). "The harvest of beluga whales in Canada's western Arctic: hunter based monitoring of the size and composition of the catch." Arctic 55(1): 10-20. Heide-Jørgensen, M. P. (2002). Narwhal (Monodon monoceros). Encyclopedia of Marine Mammals. W. F. Perrin, B. Würsig and J. G. M. Thewissen. San Diego, Academic Press: 783-787. Heide-Jørgensen, M. P. and Teilmann, J. (1994). "Growth, reproduction, age structure, and feeding habits of white whales (Delphinapterus leucas) in west Greenland waters." Meddelelser om Grønland, Bioscience 39: 195-212. Heymans, S. J. J. (2003). Revised models for Newfoundland for the time periods 1985-1987 and 1995-1997. Ecosystem Models of Newfoundland and Southeastern Labrador: Additional information and analyses for ‘Back to the Future’. S. J. J. Heymans. Vancouver, BC, Fisheries Centre Research Reports. 11(5): 40-61. Heymans, S. J. J. (2005). Ecosystem models of the western and central Aleutian Islands in 1963, 1979, and 1991. Fisheries Centre Research Report. S. Guenette and V. Christensen. Vancouver, BC, University of British Columbia. 13(1): 8-82. Higdon, J. (2007). Status of knowledge on killer whales (Orcinus orca) in the Canadian Arctic. Canadian Science Advisory Secretariat, Department of Fisheries and Oceans: 41. Higdon, J. (2008). Commercial and subsistence harvests of bowhead whales (Balaena mysticetus) in eastern Canada and west Greenland. Canadian Science Advisory Secretariat. Department of Fisheries and Oceans. Winnipeg, Canada. Higdon, J. and Ferguson, S. H. (2010). Past, present, and future for bowhead whales (Balaena mysticetus). A Little Less Arctic: Top predators in the world's largest northern inland sea, Hudson Bay. S. H. Ferguson, L. Loseto and M. Mallory. Heideberg, Germany, Springer: 159-177. Higdon, J. W. and Ferguson, S. H. (2009). "Loss of Arctic sea ice causing punctuated change in sightings of killer whales (Orcinus Orca) over the past century." Ecological Applications 19(5): 1365-1375. Hochheim, K., Barber, D. G. and Lukovich, J. V. (2010). Changing ice conditions in Hudson Bay, 1980-2005. A Little Less Arctic: Top Predators in the World's Largest Northern Inland Sea, Hudson Bay. S. H. Ferguson, L. Loseto and M. Mallory, Springer: 39-51. Holst, M., Stirling, I. and Hobson, K. A. (2001). "Diet of ringed seals (Phoca hispida) on the east and west sides of the north water polynya, northern Baffin Bay." Marine Mammals Science 17(4): 888-908. Hunt, G. L. J., Kato, H. and Mckinnell, S. M. (2000). Predation by marine birds and mammals in the subarctic North Pacific Ocean. PICES Scientific Report. N. P. M. S. Organization. Sidney, BC: 168. Hunter, A. (2005). A multiple regression model for predicting the energy requirements of marine mammals. Zoology. Vancouver, BC, University Of British Columbia. M.Sc: 300. Huntington, H. P. (2009). "A preliminary assessment of the threats to arctic marine mammals and their conservation in the coming decades." Marine Policy 33: 77-82. Jackson, G. D., Finn, J. and Nicol, S. (2002). "Planktonic cephalopods collected off East Antarctica during the 'BROKE' survey." Deep Sea Research I 49: 1049-1054. Jarre-Teichmann, A., Brey, T., Bathmann, U. V., Dahm, C., Dieckmann, G. S., Gorny, M., Klages, M., Pages, F., Plötz, J., Schnack-Schiel, S. B. and Stiller, M. (1997). Trophic flows in the benthic shelf community of the eastern Weddell Sea, Antarctica. Antarctic communities: species, structure, and survival. B. Battaglia, J. Valencia and D. W. H. Walton. Cambridge, UK, Cambridge University Press: 118-134 of 464p. JCNB (2009). Joint meeting of the NAMMCO scientific working group on the population status of narwhal and beluga in the north Atlantic and Canada/Greenland Joint commission on conservation and management of narwhal and beluga scientific working group. R. Hobbs and S. H. Ferguson. Winnipeg, MB. 17-20 Febuary, 2009. Johnson, S. R., Wiggins, D. A. and Wainwright, P. F. (1993). "Late-summer abundance and distribution of marine birds in Kesegaluk Lagoon, Chukchi Sea, Alaska." Arctic 46(3): 212-227. Karpouzi, V. S. (2005). Modelling and mapping trophic overlap between fisheries and the world's seabirds. Faculty of Graduate Studies, University of British Columbia. M.Sc.: 159. 46  An Ecosystem Model of Hudson Bay, Canada    Kastelein, R. A. (2002). Walrus (Odobenus rosmarus). Encyclopedia of Marine Mammals. W. F. Perrin, B. Würsig and J. G. M. Thewissen. San Diego, Academic Press: 1294-1300. Kelley, T. C., Loseto, L. L., Stewart, R. E. A., Yurkowski, M. and Ferguson, S. H. (2010). Importance of eating capelin: Unique dietary habits of Hudson Bay beluga. A Little Less Arctic: Top predators in the World's Largest Northern Inland Sea, Hudson Bay. S. H. Ferguson, L. L. Loseto and M. M.L. London, Springer: 53-69. Kovacs, K. M. (2002). Bearded Seal (Erignathus barbatus). The Encyclopedia of Marine Mammals. W. F. Perrin, B. Würsig and J. G. M. Thewissen. San Diego, Academic Press: 84-87. Lavigne, D. M. (1979). Harp Seal. Mammals in the Seas, volume II: pinniped species summaries and report on sirenians. FAO Fisheries Series No. 5. Rome: 76-80. Lavigne, D. M. (2002). Harp seal (Pagophilus groenlandicus). The Encyclopedia of Marine Mammals. W. F. Perrin, B. Würsig and J. G. M. Thewissen. San Diego, Academic Press: 560-562. Lawson, J., Stevens, T. and Snow, D. (2007). Killer whales of Atlantic Canada, with particular reference to the Newfoundland and Labrador region. Canadian Science Advisory Secretariat. D. o. F. a. Oceans: 1-16. Lee, J. and Taylor, M. (1994). Aspects of the polar bear harvest in the Northwest territories, Canada. Bears: Their Biology and Management, Missoula, Montana. Legendre, L., Robineau, B., Gosselin, M., Michel, C., Ingram, R. G., Fortier, L., Therriault, J. C., Demers, S. and Monti, D. (1996). "Impact of freshwater on a subarctic coastal ecosystem under seasonal sea ice (southeastern Hudson Bay) Canada II. Production and export of microalgae." Journal of Marine Systems 7: 233-250. Loseto, L. L., Stern, G. A., Connelly, T. L., Deibel, D., Gemmill, B., Prokopowicz, A., Fortier, L. and Ferguson, S. H. (2009). "Summer diet of beluga whales inferred by fatty acid analysis of the eastern Beaufort Sea food web." Journal of Experimental Marine Biology and Ecology 374: 12-18. Lowry, L. F., Frost, K. J. and Burns, J. J. (1987). "Food of ringed seals and bowhead whales near Point Barrow Alaska." The Canadian Field Naturalist 92: 67-70. Lunn, N. J., Atkinson, S., Branigan, M., Calvert, W., Clark, D., Doidge, B., Elliot, C., Nadgy, J., Obbard, M., Otto, R., Stirling, I., Taylor, M., Vandal, D. and Wheatley, M. (2002). Polar bear management in Canada 1997-2000. Polar Bears. Proceedings of the 13th Working meeting of the ICUN/SSC Polar Bear Specialist Group 23-38 June 2001, Nuuk, Greenland. Occas. Pap. ICUN species Surviv. Comm. . N. J. Lunn, S. Schliebe and E. D. Born: 26. Lunn, N. J., Stirling, I. and Nowicki, S. N. (1997). "Distribution and abundance of ringed (Phoca hispida) and bearded seals (Erignathus barbatus) in western Hudson Bay." Canadian Journal of Fisheries and Aquatic Sciences 54: 914-921. Luque, S. P. and Ferguson, S. H. (2008). Movement and behaviour of Hudson Bay ringed seals (Phoca hispida). Poster presentation Arctic Change conference. Quebec, Canada. Luque, S. P. and Ferguson, S. H. (2010). "Age structure, growth, mortality, and density of belugas (Delphinapterus leucas) in the Canadian Arctic: responses to the environment?" Polar Biology 33: 163-178. Mansfield, A. W. (1967a). "Distribution of the harbor seal, Phoca vitulina Linnaeus, in Canadian Arctic waters." Journal of Mammalogy 48(2): 249-257. Mansfield, A. W. (1967b). "Seals of Arctic and eastern Canada." Fisheries Research Board of Canada Bulletin 137: 1-35. Mansfield, A. W. and St Aubin, D. J. (1991). "Distribution and abundance of the Atlantic walrus, Odobenus rosmarus rosmarus, in the Southampton Island-Coats Island region of Northern Hudson Bay." Canadian Field Naturalist 105: 95-100. Metz, C. and Schnack-Schiel, S. B. (1995). "Observations on carnivorous feeding in Antarctic calanoid copepods." Marine Ecology Progress Series 129: 71-75. Migratory Birds Convention Act (1994). c.22 June 3, 1994. Department of Justice Canada. Mikhail, M. Y. and Welch, H. E. (1989). "Biology of Greenland cod, Gadus ogac, at Saqvaqjuac, northwest coast of Hudson Bay." Environmental Biology of Fishes 26: 49-62. Miyazaki, N. (2002). Ringed, Caspian, and Baikal seals. The Encyclopedia of Marine Mammals. W. F. Perrin, B. Würsig and J. G. M. Thewissen. San Diego, Academic Press: 1033-1037. Muir, D. C. G., Born, E. W., Koczansky, K. and Stern, G. (2000). "Temporal and spatial trends of persistant organochlorines in Greenland walrus (Odobenus rosmarus rosmarus)." The Science of the Total Environment 245: 73-86. Muir, D. C. G., Segstro, M. D., Hobson, K. A., Ford, C. A., Stewart, R. E. A. and Olpinski, S. (1995). "Can seal eating explain elevated levels of PCB's and organochlorine pesticides in walrus blubber from eastern Hudson Bay (Canada)?" Environmental Pollution 90(3): 335-348. NAMMCO (2005a). The Atlantic walrus. Status of Marine Mammals in the North Atlantic. North Atlantic Marine Mammal Commission. Tromso, Polar Environmental Centre: 1-7. NAMMCO (2005b). The Beluga Whale. Status of marine mammals in the North Atlantic. North Atlantic Marine Mammal Commmission. Tromso: 11. From the Tropics to the Poles, Wabnitz and Hoover  Nicklen, P. (2007). "Arctic Ivory." National Geographic(August 2007). Nunavut Bureau of Statistics (2008). "Nunavut population counts by community, 1981 Census to 2006 Census." Retrieved Sept 15,2009, from http://www.gov.nu.ca/eia/stats/census.html. Okey, T. A. and Pauly, D. (1999). 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. T. Okey and D. Pauly. Vancouver, BC. 7(4): 1-136. Paetkau, D., Amstrup, S. C., Born, E. W., Calvert, W., Derocher, A. E., Garner, G. W., Messier, F., Stirling, I., Taylor, M. K., Wiig, Ø. and Strobeck, C. (1999). "Genetic structure of the world's polar bear populations." Molecular Ecology 8: 1571-1584. Pakhomov, E. A., Perissinotto, R., Froneman, P. W. and Miller, D. G. M. (1997). "Energentics and feeding dynamics of Euphausia superba in the South Georgia region during the summer of 1994." Journal of Plankton Research 19(4): 399-423. Palomares, M. L. D. and Pauly, D. (1998). "Predicting food consumption of fish populations as functions of mortality, food type, morphometrics, temperature and salinity." Marine and Fisheries Research 49: 447-453. Parkinson, C. L., Cavalieri, D. J., Gloerson, P., Zwally, J. H. and Cosimo, J. C. (1999). "Arctic sea ice extents, areas, and trends 1978-1996." Journal of Geophysical Research Oceans 104(C9 20): 837-856. Pauly, D. (1980). "On the interrelationships between natural mortality, growth parameters, and mean environmental temperature in 175 fish stocks." Journal du Conseil ICES/CIEM 39(2): 175-192. Pauly, D. and Christensen, V. (1995). "Primary production required to sustain global fisheries." Nature 374: 255-257. Pauly, D., Sambilay, V. J. and Opitz, S. (1993). Estimates of relative food consumption by fish and invertebrate populations, required for modelling the Boliano reef ecosystem, Philippines. Trophic models of aquatic ecosystems. V. Christensen and D. Pauly, ICLARM Conference Proceedings 26, 390p. 26: 236-251. Pauly, D., Trites, A. W., Capuli, E. and Christensen, V. (1998). "Diet composition and trophic levels of marine mammals." ICES Journal of Marine Science 55: 467-481. Peacock, E., Derocher, A. E., Lunn, N. J. and Obbard, M. E., Eds. (2010). Polar bear ecology and management in Hudson Bay in the face of climate change. A Little Less Arctic: Top predators in the world's largest northern inland sea: Hudson Bay. Heidelberg, Germany, Springer. Peterson, S., Bortoluzzi, T. and Ferguson, S. H. (2009). Killer whale update for northerners. Aarluk News. Winnipeg, MB, University of Manitoba. Issue 1. Poltermann, M. (2001). "Arctic sea ice as a feeding ground for amphipods- food sources and strategies." Polar Biology 24: 89-96. Por, F. D. and Bromley, H. J. (1974). "Morphology and anatomy of Maccabeus tentaculatus (Priapulida: Seticoronaria)." Journal of Zoology 173(2): 173-197. Pruvost, P., Duhamel, G. and Palomares, M. L. D. (2005). An ecosystem model of the Kerguelen Islands EEZ. Fisheries Centre Research Reports. M. L. D. Palomares, P. Pruvost, T. J. Pitcher and D. Pauly, University of British Columbia, Vancouver, Canada. 13(7): 40-64. Rayner, N. A., Parker, D. E., Horton, E. B., Folland, C. K., Alexander, L. V., Rowell, D. P., Kent, E. C. and Kaplan, A. (2003). "Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century." Journal of Geophysical Research 108(D14): 29. Richard, P. (1991). "Abundance and distribution of Narwhals (Monodon monoceros) in Northern Hudson Bay." Canadian Journal of Fisheries and Aquatic Sciences 48: 276-283. Richard, P. (2005). An estimate of the western Hudson Bay beluga population size in 2004. Canadian Science Advisory Secretariat. Department of Fisheries and Oceans Canada. Winnipeg, Canada: i+29. Richard, P. R. and Campbell, R. R. (1988). "Status of the Atlantic walrus Odobenus rosmarus rosmarus, in Canada." Canadian Field Naturalist 102: 337-350. Riebesell, U., Schloss, I. and Smetacek, V. (1991). "Aggregation of algae released from melting sea ice: implications for seeding and sedimentation." Polar Biology 11: 239-248. Roberge, M. M. and Dunn, J. B. (1990). Assessment of the subsistence harvest and biology of narwhal (Monodon monoceros L.) from Admiralty Inlet, Baffin Island, N.W.T., 1983 and 1986-89. Canadian Technical Report of Fisheries and Aquatic Sciences, Depart of Fisheries and Oceans. N. 1747: 1-39. Rodhouse, P. G. and White, M. G. (1995). "Cephalopods occupy the ecological niche of the eppipelagic fish in the Antarctic polar frontal zone." Biological Bulletin 189: 77-80. Rugh, D. J. and Shelden, K. E. (2002). Bowhead whale (Balaena mysticetus). Encyclopedia of marine mammals. W. F. Perrin, B. Würsig and J. G. M. Thewissen. San Diego, Academic Press: 129-131. Runge, J. A. and Ingram, R. G. (1987). "Underice grazing by planktonic, calanoid copepods in relation to a bloom of ice microalgae in southeastern Hudson Bay." Limnology and Oceanography 33(2): 280-286. 48  An Ecosystem Model of Hudson Bay, Canada    Runge, J. A. and Ingram, R. G. (1991). "Under-ice feeding and diel migration by the planktonic copepods Calanus glacialis and Pseudocalanus minutus in relation to the ice algal production cycle in southeastern Hudson Bay, Canada." Marine Biology 108: 217-225. Russell, R. H. (1975). "The food habits of polar bears of James Bay and southwest Hudson Bay in summer and autumn." Arctic 28(2): 117-129. Smith, C. R., Mincks, S. and Demaster, D. J. (2006). "A synthesis of bentho-pelagic coupling on the Antarctic shelf: Food banks, ecosystem inertia and global climate change." Deep Sea Research II 53: 875-894. Smith, R. J., Hobson, K. A., Koopman, H. A. and Lavigne, D. M. (1996). "DIstinguishing between populations of fresh-and salt-water harbor seals (Phoca vitulina) using stable-isotope ratios and fatty acid signatures." Canadian Journal of Fisheries and Aquatic Sciences 53: 272-279. Smith, R. J. and Lavigne, D. M. (1994). "Subspecific status of the freshwater harbour seal (Phoca vitulina mellonae) a re-assessment." Marine Mammal Science 10(1): 105-110. Smith, T. G. (1975). "Ringed seals in James Bay and Budson Bay: Population estimates and catch statistics." Arctic 28: 170-182. Smith, T. G. (1981). Notes on the bearded seal, Erignathus barbatus, in the Canadian arctic. Canadian Technical Report of Fisheries and Aquatic Sciences Government of Canada Fisheries and Oceans. 1042: v+49. Statistics Canada (2006). Census Canada. Stewart, D. B. and Lockhart, W. L. (2005). An overview of the Hudson Bay marine ecosystem. Canadian technical report of fisheries and aquatic sciences; 2586. Arctic Biological Consultants; Canada Department of Fisheries and Oceans Central and Arctic Region. Winnipeg, MB: 487. Stewart, R. E. A. (2008). "Redefining walrus stocks in Canada." Arctic 61(3): 292-308. Stewart, R. E. A., Campana, S. E., Jones, C. M. and Stewart, B. E. (2006). "Bomb radiocarbon dating calibrates beluga (Delphinapterus leucas) age estimates." Canadian Journal of Zoology 18: 1840-1852. Stirling, I. (1997). "The importance of polynyas, ice edges, and leads to marine mammals and birds." Journal of Marine Systems 10: 9-21. Stirling, I. (2002). Polar Bear (Urus maritimus). Encyclopedia of marine mammals. W. F. Perrin, B. Würsig and J. G. M. Thewissen. San Diego, Academic Press: 945-948. Stirling, I. and Derocher, A. E. (1993). "Possible impacts of climatic warming on polar bears." Arctic 46(3): 240-245. Stirling, I., Lunn, N. J. and Iacozza, J. (1999). "Long-term trends in population ecology of polar bears in Western Hudson Bay in relation to climate change." Arctic 52(3): 294-306. Stirling, I. and Parkinson, C. L. (2006). "Possible effects of climate warming on selected populations of polar bears (Ursus maritimus) in the Canadian Arctic." Arctic 59(3): 261-275. Straneo, F. and Saucier, F. (2008). "THe outflow from Hudson Strait and its contribution to the Labrador Current." Deep Sea Research I 55: 926-946.  Strong, J. T. (1989). Reported harvests of narwhal, beluga, and walrus in the Northwest Territories, 1948-1987. Canadian Data Report of Fisheries and Aquatic Sciences. Department of Fisheries and Oceans Canada, Central and Arctic Region: 14.  Sutherland, T., Chakasim, T. and Lantz, P. (2010). "The Moose Cree First Nations Website." Community Profile. from http://www.moosecree.com/our_community/profile.html.  Tönnesson, K. and Tiselius, P. (2005). "Diet of the chaetognaths Sagitta setosa and S. elegans in relation to prey abundance and vertical distribution." Marine Ecology Progress Series 289: 177-190.  Tremblay, C., Runge, J. A. and Legendre, L. (1989). "Grazing and sedimentation of ice algae during and immediately after a bloom at the ice-water interface." Marine Ecology Progress Series 56: 291-300.  Trites, A. and Pauly, D. (1998). "Estimating mean body mass of marine mammals from measurements of maximum body lengths." Canadian Journal of Zoology 76: 886-896.  Walters, C., Christensen, V. and Pauly, D. (1997). "Structuring dynamic models of exploited ecosystems from trophicmass-balance assessments." Reviews in Fish Biology and Fisheries 7(2): 139-172. Weaver, P. A. and Walker, R. S. (1988). The narwhal (Monodon monoceros L.) harvest in Pond Inlet, Northwest Territories: Hunt documentation and biological sampling, 1982-1983. Canadian Manuscript Report of Fisheries and Aquatic Sciences. Department of Fisheries and Oceans. No. 1975: 26. Westdal, K. H., Richard, P. R. and Orr, J. R. (2010). Migration routes and seasonal home range of the northern Hudson Bay narwhal (Monodon monoceros). A little less Arctic: Top predators in the world's largest northern inland sea, Hudson Bay. S. H. Ferguson, L. Loseto and M. Mallory. London, UK, Springer: 71-92. Young, B. G., Loseto, L. L. and Ferguson, S. H. (2010). "Diet differences among age classes of Arctic seals: evidence from stable isotope and mercury biomarkers." Polar Biology 33: 153-162.  From the Tropics to the Poles, Wabnitz and Hoover  APPENDIX 1 : MARINE MAMMAL MORTALITY CALCULATIONS Mortality for marine mammal functional groups was calculated based on life history information and estimates of longevity, using equation 1 to estimate the probability of survivorship from birth to age x, with information from equations 2-4, and parameters in table A1.   ݁ݍ. ሺ1ሻ																																				ܮሺݔሻ ൌ ܮ௝ሺݔሻ ∙ ܮ௖ሺݔሻ ∙ ܮ௦	ሺݔሻ ݁ݍ. ሺ2ሻ																																				ܮ௝ሺݔሻ ൌ ݁ݔ݌ሾሺെܽଵ ܾଵ⁄ ሻ ∙ ሼ1 െ ݁ݔ݌ሺെܾ ∙ ݔ ߗ⁄ ሻሽሿ				 	݁ݍ. ሺ3ሻ																																			ܮ௖ሺݔሻ ൌ 	݁ݔ݌ሾെܽଶ ∙ ݔ ߗ⁄ ሿ				 ݁ݍ. ሺ4ሻ																																				ܮ௦ሺݔሻ ൌ ݁ݔ݌ሾሺܽଷ ܾଷ⁄ ሻ ∙ ሼ1 െ ݁ݔ݌ሺܾଷ ∙ ݔ ߗ⁄ ሻሽሿ  Table A1: Parameters for equations 1-4 	ܮ௝ሺݔሻ Mortality due to juvenile factors 	ܮ௖ሺݔሻ Constant mortality experienced by all age classes 	ܮ௦ሺݔሻ Mortality due to senescent factors ܽଵ, ܽଶ,ܽଷ, ܾଵ, ܾଷ Allow flexibility in shape of survivorship curve ߗ Longevity  For all pinniped groups survivorship curve parameters from northern fur seals were used to estimate survivorship (table A2). Human survivorship parameter were used for killer whales, as there are few to zero predators on this group, likely causing lowered juvenile mortality. Baleen whale (bowhead whales) and beluga whale survivorship was calculated using monkey and human survivorship parameters, however the monkey parameters were used as they had a slightly higher juvenile mortality. This was believed to be more representative of baleen whale survivorship. Mortality was calculated as 1- the survivorship for each year of longevity, and averaged over all ages (x) to give the P/B value.   Table A2: Survivorship curve parameters based on life histories of fur seals, monkeys, and humans. Species group ࢇ૚ ࢇ૛ ࢇ૜ ࢈૚ ࢈૜ Northern Fur Seal 14.343 0.1710 0.0121 10.259 6.6878 Old World Monkeys 30.430 0.0000 0.7276 206.720 2.3188 Human (female) 40.409 0.4772 0.0047 310.360 8.0290   50  An Ecosystem Model of Hudson Bay, Canada    APPENDIX 2 : BIRD SPECIES FOUND WITHIN THE HUDSON BAY MODEL AREA BY FAMILY, AS REPORTED FROM STEWART AND LOCKHART (2005).  Family Gaviidae: Loons red-throated loon    Gavia stellata (Pontoppidan, 1763)  Pacific loon    G. pacifica (Lawrence)  common loon     G. immer (Brünnich)  yellow-billed loon    G. adamsii (Gray) *  Family Podicipedidae: Grebes pied-billed grebe    Podilymbus podiceps (Linnaeus) * horned grebe     Podiceps auritus (Linnaeus)  Family Procellariidae: Fulmars northern fulmar    Fulmarus glacialis (Linnaeus) *  Family Hydrobatidae: Storm-petrels Leach's storm-petrel    Oceanodroma leucorhoa (Viellot) *  Family Pelecanidae: Pelicans American white pelican    Pelecanus erythrorhynchos (Gmelin) *  Family Sulidae: Gannets northern gannet    Sula bassanus (Linnaeus) *  Family Phalacrocoracidae: Cormorants double crested cormorant    Phalacrocorax auritus (Lesson)   Family Ardeidae: Herons and Bitterns American bittern    Botaurus lentiginosus (Rackett)  great blue heron    Ardea herodias (Linnaeus)  snowy egret     Egretta thula (Molina) * little blue heron    E. caerulea (Linnaeus) * tricolor heron    E. tricolor (Müller) * black-crowned night heron   Nycticorax nycticorax (Linnaeus) *  Family Anatidae: Geese, Swans, and Ducks greater white-fronted goose   Anser albifrons (Scopoli)  snow goose       Chen caerulescens (Linnaeus)  Ross's goose     C. rossii (Cassin)  Canada goose     Branta canadensis (Linnaeus)  Brant      B. bernicla (Linnaeus)  trumpeter swan    Cygnus buccinator (Richardson) * tundra swan     C. columbianus (Ord)  gadwall     Anas strepera (Linnaeus) * Eurasian widgeon    A. penelope  (Linnaeus) * American widgeon (baldpate)   A. americana (Gmelin)  American black duck    A. rubripes (Brewster)  mallard     A. platyrhynchos (Linnaeus)  blue winged tea l    A. discors (Linnaeus) northern shoveler    A. souchet (Linnaeus)  northern pintail    A. acuta (Linnaeus)  green-winged teal    A. crecca (Linnaeus)  canvasback     Aythya valisineria (Wilson) * From the Tropics to the Poles, Wabnitz and Hoover  APPENDIX 2 (CONTINUED) : BIRD SPECIES FOUND WITHIN THE HUDSON BAY MODEL AREA Family Anatidae: Geese, Swans, and Ducks (Continued) redhead     A. americana (Eyton) * ring-necked duck    A. collaris (Donovan)  greater scaup     A. marila (Linnaeus)  lesser scaup     A. affinis (Eyton)  king eider     Somateria spectabilis (Linnaeus)  common eider       S. mollissima (Linnaeus)  harlequin ducks     Histrionicus histrionicus (Linnaeus)  surf scoter     Melanitta perspicillata (Linnaeus) white-winged scoter    M. fusca (Linnaeus)  black scoter (common scoter)   M. nigra (Linnaeus)  long-tailed duck (oldsquaw)   Clangula hyemalis (Linnaeus)  bufflehead     Bucephala albeola (Linnaeus) * common goldeneye    B. clangula (Linnaeus)  Barrow's goldeneye    B. islandica (Gmelin) * hooded merganser    Lophodytes cucullatus (Linnaeus) * common merganser    Mergus merganser Linnaeus  red-breasted merganser    M. serrator (Linnaeus)  ruddy duck     Oxyura jamaicensis (Gmelin) *  Family Accipiteridae: Ospreys, Eagles, Hawks, and Allies osprey      Pandion haliaetus (Linnaeus)  bald eagle     Haliaeetus leucocephalus (Linnaeus)  northern harrier (marsh hawk)   Circus cyaneus (Linnaeus)  northern goshawk    Accipter gentilis (Wilson) * sharp-shinned hawk    A. striatus (Vieillot)  rough-legged hawk    Buteo lapopus (Gmelin)  golden eagle     Aquila chrysaetos (Linnaeus) *  Family Falconidae: Falcons merlin      Falco columbarius (Linnaeus) peregrine falcon      F. peregrinus (Tunstall)  gyrfalcon     F. rusticolus (Linnaeus) prairie falcon     F. mexicanus (Schlegel) *  Family Rallidae: Rails, Gallinules, and Coots yellow rail      Coturnicops noveboracensis (Gmelin) sora      Porzana carolina (Linnaeus)  American coot     Fulica americana Gmelin   Family Gruidae: Cranes sandhill crane     Grus canadensis (Linnaeus)   Family Charadriidae: Plovers black-bellied plover    Pluvialis squatarola (Linnaeus) American golden-plover   P. dominica (Muller) semipalmated plover    Charadrius semipalmatus (Bonaparte) killdeer     C. vociferus (Linnaeus)   Family Scolopacidae: Sandpipers, Phalaropes, and allies greater yellowlegs    Tringa melanoleuca (Gmelin) lesser yellowlegs    T. flavipes (Gmelin)  solitary sandpiper    T. solitaire (Wilson) spotted sandpiper    Actitis macularia (Linnaeus) 52  An Ecosystem Model of Hudson Bay, Canada    APPENDIX 2 (CONTINUED) : BIRD SPECIES FOUND WITHIN THE HUDSON BAY MODEL AREA   Family Scolopacidae: Sandpipers, Phalaropes, and allies (Continued) whimbrel      Numenius phaeopus (Linnaeus)  Hudsonian godwit      Limosa haemastica (Linnaeus) marbled godwit    L. fedoa (Linnaeus) ruddy turnstone    Arenaria interpres (Linnaeus) red knot       Calidris canutus (Linnaeus) sanderling     C. alba (Pallas)  semipalmated sandpiper   C. pusilla (Linnaeus) little stint     C. minuta (Leisler)* least sandpiper     C. minutilla (Vieillot)  white-rumped sandpiper   C. fuscicollis (Vieillot)  Baird's sandpiper    C. bairdii (Coues)  pectoral sandpiper    C. melanotos (Vieillot)  purple sandpiper    C. maritima (Brunnich) dunlin      C. alpina (Linnaeus) stilt sandpiper     C. himantopus (Bonaparte) buff-breasted sandpiper   Tryngites subruficollis (Vieillot)  short-billed dowitcher    Limnodromus griseus (Gmelin)  Wilson's snipe     Gallinago delicata (Ord)  Wilson's phalarope    Phalaropus tricolor (Vieillot)  red-necked/northern phalarope   P. lobatus (Linnaeus)  red phalarope     P. fulicaria (Linnaeus)   Family Laridae: Jaegers, Gulls, and Terns Pomeranian jaeger    Stercorarius pomarinus (Temminick)  parasitic jaeger     S. parasiticus (Linnaeus) long-tailed jaeger    S. longicaudus (Vieillot) laughing gull     Larus atricilla (Linnaeus) * Franklin's gull     L. pixican (Wagler)* little gull     Larus minutus (Pallas) black-headed gull    L. ridibundus (Linnaeus) * Bonaparte's gull    L. philadelphia (Ord)  mew gull     L. canus (Linnaeus) * ring-billed gull     L. delawarensis (Ord)  California gull     L. californicus (Lawrence) * herring gull     L. argentatus (Pontoppidan)  Iceland gull     L. glaucoides (Meyer)  lesser black-backed gull   L. fuscus (Linnaeus) * glaucous -winged gull    L. glaucescens (Naumann) * glaucous gull     L. hyperboreus (Gunnerus)  great black-backed gull    L. marinus (Linnaeus) * black-legged kittiwake    Rissa tridactyle (Linnaeus)  Ross's gull11     Rodostethia rosea (MacGillivray)  Sabine's gull     Xema sabini (Sabine) ivory gull12     Pagophila eburnea (Phipps) * Caspian tern     Sterna caspia (Pallas) common tern     S. hirundo (Linnaeus)  Arctic tern     S. parasisaea (Pontoppidan)  Forster's tern     S. forsteri (Nuttall) * white-winged tern    Chlidonias leucopterus (Temminck) * black tern     C. niger (Linnaeus)    From the Tropics to the Poles, Wabnitz and Hoover  APPENDIX 2 (CONTINUED) : BIRD SPECIES FOUND WITHIN THE HUDSON BAY MODEL AREA   Family Alcidae: Auks, Murres, and Puffins Dovekie     Alle alle (Linnaeus)  thick-billed murre    Uria lomvia (Linnaeus) black guillemot     Cepphus grylle (Linnaeus)  Family Strigidae: Typical owls snowy owl     Nyctea scandiaca (Linnaeus)  short-eared owl      Asio fla meus (Pontoppidan)   Family Alcedinidae: Kingfishers belted kingfisher    Ceryle alcyon (Linnaeus)  Family Corvidae: Crows and Ravens American crow     Corvus brachyrhynchos (Brehm)  common raven     C. corax (Linnaeus)   Family Alaudidae: Larks horned lark     Eremophila alpestris   Family Motacillidae: Pipits American pipit     Anthus rubescens (Tunstall)   *Indicates the species is rare in its distribution within the model area as classified by Stewart and Lockhart (2005).                             54  An Ecosystem Model of Hudson Bay, Canada    APPENDIX 3:  FISH FUNCTIONAL GROUPS AND SPECIES INCLUDED IN EACH GROUPArctic Char:  Salvelinus alpinus Arctic Char Atlantic Salmon: Salmo salar Atlantic Salmon Gadiformes: Arctogadus glacialis polar cod Boreogadus saida Arctic cod Gadus ogac Greenland cod Sculpins/ Zoarcids:  Gymnocanthus tricuspis Arctic staghorn Icelus bicornis twohorn sculpin Icelus spatula spatulate sculpin Myoxocephalus quadricornis fourhorn sculpin Myoxocephalus scorpioides Arctic sculpin Myoxocephalus scorpius shorthorn sculpin Triglops murrayi moustache sculpin Triglops pingelli ribbed sculpin Gymnelus viridis fish doctor Lycodes pallidus pale eelpout Lycodes reticulatus Arctic eelpout Other Marine Fish: Stichaeus punctatus Arctic shanny Lumpenus fabricii slender eelblenny Pleuronectidae sp. righteye flounder Hippoglossoides platessoides Canadian plaice Capelin: Mallotus villosus Capelin  Sandlance: Ammodytes dubius northern sand lance Ammodytes hexapterus stout sand lance Sharks/Rays: Somniosidae sleeper sharks Rajidae skates    Other Marine Fish: Leptagonus decagonus alligator poacher Ulcina olriki Atlantic alligatorfish Cyclopterus lumpus lumpfish Eumicrotremus derjugini leatherfin lumpsucker Eumicrotremus spinosus Atlantic spiny lumpsucker Careproctus reinhardti sea tadpole Liparis fabricii gelatinous snailfish Liparis gibbus dusky snailfish Liparis tunicatus kelp snailfish Anisarchus medius stout eelblenny Eumesogrammus praecisus fourline snakeblenny Leptoclinus maculatus daubed shanny Pholis fasciata banded gunnel Clupea harengus Atlantic Herring    From the Tropics to the Poles, Wabnitz and Hoover  APPENDIX 4: VULNERABILITIES USED TO FIT THE MODEL      Prey \ predator 1 2 3 4 5 6 7 8 9 10 11 12 13 1 Polar Bear WHB 2 2 SH Polar Bear 3 3 Polar Bear Foxe 2 4 Killer Whale 5 Narwhal 10 6 Bowhead 10 7 Walrus N 2 10 8 Walrus S 3 10 9 Bearded Seal 2 3 2 10 2 10 Harbour Seal 2 3 2 10 11 Ringed Seal 2 3 2 10 10 12 Harp seal 2 3 2 10 13 Beluga E 3 2 14 Beluga W 2 2 2 15 Beluga James 10 2 16 Seabirds 2 3 2 10 17 Arctic Char 1 1 18 Atlantic Salmon 10 1 1 1 2 19 Gadiformes 10 1 10 2 2 10 10 2 1 20 Sculpins/ Zoarcids 10 1 2 2 1 10 10 1 1 21 Capelin 1 10 2 1 1 2 22 Sandlance 10 2 1 23 Sharks/Rays 2 24 Other Marine Fish 2 10 1 10 10 1 2 2 1 25 Brackish Fish 2 1 1 2 2 26 Cephalopods 2 10 1 1 2 2 2 27 MacroZooplankton 2 1 2 1 2 2 28 Euphausids 2 1 2 2 2 2 29 Copepods 2 2 2 30 Crustaceans 2 1 2 10 10 1 2 2 1 2 31 Other MesoZooplankton 2 2 32 MicroZooplankton 2 2 33 Marine Worms 2 2 10 10 10 10 10 34 Echinoderms 10 10 2 2 10 10 10 10 35 Bivalves 10 10 2 2 10 10 36 Other Benthos 10 2 2 10 10 10 10 10 10 37 Primary Production 2 2 2 38 Ice Algae 2 39 Ice Detritus 2 40 Pelagic Detritus 2 2 56  An Ecosystem Model of Hudson Bay, Canada    APPENDIX 4 (CONTINUED): VULNERABILITIES USED TO FIT THE MODEL Prey \ predator 14 15 16 17 19 20 21 22 23 24 25 26 1 Polar Bear WHB     2 SH Polar Bear     3 Polar Bear Foxe     4 Killer Whale     5 Narwhal     2 6 Bowhead     7 Walrus N     8 Walrus S     9 Bearded Seal     2 10 Harbour Seal     11 Ringed Seal     2 12 Harp seal     2 13 Beluga E     14 Beluga W     15 Beluga James     16 Seabirds   2  17 Arctic Char 2  2  2 2 18 Atlantic Salmon 2 2 2 1 2 2 19 Gadiformes 10 2 2 2 2 1 10 20 Sculpins/ Zoarcids 2  2 2 2 1 10 21 Capelin 2 1 2 1 10 10 10 1 2 1 22 Sandlance 2  2 1 10 10 10 2 1 23 Sharks/Rays     2 24 Other Marine Fish 2  2 1 10 10 2 2 25 Brackish Fish 2  2 1 2 1 26 Cephalopods 2 1 2  1 2 1 1 2 27 MacroZooplankton 2  2 1 1 1 2 1 1 1 28 Euphausids 2 2 2 1 2 2 2 1 1 1 29 Copepods 2 2 2 1 2 2 1 1 1 30 Crustaceans 2 2  1 2 2 2 2 2 1 1 1 31 Other MesoZooplankton   2 1 2 2 1 1 1 32 MicroZooplankton    1 2 2 1 1 1 33 Marine Worms 10 10 10 10 10 10 10 10 10 34 Echinoderms   10 10 10 10 10 10 35 Bivalves   10  10 10 10 10 36 Other Benthos 10 10 10 10 10 10 10 10 10 37 Primary Production    1 1 1 1 1 38 Ice Algae    1 2 2 1 1 39 Ice Detritus     2 2 40 Pelagic Detritus   1  1 1 1 1     From the Tropics to the Poles, Wabnitz and Hoover  APPENDIX 4 (CONTINUED): VULNERABILITIES USED TO FIT THE MODEL Prey \ predator 27 28 29 30 31 32 33 34 35 36 1 Polar Bear WHB           2 SH Polar Bear           3 Polar Bear Foxe           4 Killer Whale           5 Narwhal           6 Bowhead           7 Walrus N           8 Walrus S           9 Bearded Seal           10 Harbour Seal           11 Ringed Seal           12 Harp seal           13 Beluga E           14 Beluga W           15 Beluga James           16 Seabirds           17 Arctic Char           18 Atlantic Salmon           19 Gadiformes           20 Sculpins/ Zoarcids           21 Capelin           22 Sandlance           23 Sharks/Rays           24 Other Marine Fish           25 Brackish Fish           26 Cephalopods           27 MacroZooplankton  1     1   1 28 Euphausids 2 1  2 1  2 2   29 Copepods 2 1  1 1  1 2 1  30 Crustaceans 2 1  1 2  2 2   31 Other MesoZooplankton 2 1  2 1  1 2 1 1 32 MicroZooplankton 2 1 2 2 1  1 2 1 1 33 Marine Worms    10   10 10  10 34 Echinoderms    10   10 10  10 35 Bivalves    10    10  10 36 Other Benthos    10   10 10  10 37 Primary Production 1 1 1 1 1 1 1 1 1 1 38 Ice Algae 1 1 1 1 1 1 10 10 10 10 39 Ice Detritus    2   2 2 2 2 40 Pelagic Detritus 1 1 1 1 1 1          58  An Ecosystem Model of Hudson Bay, Canada    APPENDIX 5: MIXED TROPHIC IMPACT RESULTS FROM THE BALANCED ECOPATH MODEL  Impacting / Impacted Polar Bear WHB SH Polar Bear Polar Bear Foxe Killer Whale Narwhal Bowhead Walrus NWalrus  S Bearded Seal Harbour Seal Ringed Seal 1 Polar Bear WHB -0.53 -0.0237 -0.0328 -0.0255 0.0138 0.0035 -0.0133 0.0095 -0.0635 0.0009 -0.0607 2 SH Polar Bear -0.0123 -0.52 -0.0152 -0.017 0.005 0.0024 0.0007 -0.094 -0.0648 -0.0163 -0.0228 3 Polar Bear Foxe -0.0138 -0.0111 -0.515 -0.0123 0.0057 0.0017 0.0004 0.0045 -0.0528 -0.184 -0.0226 4 Killer Whale 0.0047 0.0052 0.0036 -0.508 -0.063 -0.0726 -0.0109 -0.0937 -0.0161 -0.103 0.026 5 Narwhal -0.0006 -0.0005 -0.0007 0.021 -0.442 -0.0031 -0.0002 -0.0043 -0.0031 -0.0053 -0.0007 6 Bowhead 8E-06 2E-05 -7E-06 0.007 -0.001 -0.427 -0.0002 -0.0013 -0.0003 -0.0016 0.0002 7 Walrus N 0.0012 -0.0002 -0.0003 0.0073 -0.0021 -0.0012 -0.477 -0.0037 -0.0017 -0.0023 0.0002 8 Walrus S -0.0593 -0.0585 -0.0543 -0.0119 0.0179 0.0009 0.0004 -0.294 0.0513 0.0931 -0.212 9 Bearded Seal 0.0417 0.042 0.0752 0.0464 -0.0057 -0.007 -0.0021 -0.0177 -0.105 -0.0922 -0.0385 10 Harbour Seal -0.0007 -0.0007 0.0037 0.0063 -0.0022 -0.0009 -0.0001 -0.0013 -0.0025 -0.0057 -0.003 11 Ringed Seal 0.194 0.197 0.172 0.067 -0.0671 -0.007 -0.0035 -0.041 -0.314 -0.309 -0.276 12 Harp seal 0.0113 0.0112 0.0157 0.0114 -0.0018 -0.0016 -0.0007 -0.0046 -0.0137 -0.0158 -0.0068 13 Beluga E -0.0003 0.001 -0.0004 0.0021 -0.0016 -0.0004 0.0001 -0.0006 -0.0011 -0.0009 -0.0006 14 Beluga W 0.0282 -0.0095 0.0194 0.0481 -0.0267 -0.0073 -0.0008 -0.0107 -0.0342 -0.0306 -0.0172 15 Beluga James -0.0013 0.0255 -0.0015 0.0189 -0.0033 -0.0028 -0.0004 -0.0088 -0.0053 -0.0062 -0.0018 16 Seabirds -0.0019 0.0102 -0.0025 -0.0018 -0.0149 0.0007 -0.0044 -0.0073 -0.017 -0.0259 -0.0108 17 Arctic Char -0.0009 -0.0024 -0.0004 0.0008 -0.0018 -0.0006 -1E-04 -0.0012 0.0195 -0.0138 -0.0115 18 Atlantic Salmon -0.0058 -0.006 -0.0057 -0.0039 -0.0167 0.0003 0.0024 0.0019 3E-06 -0.0221 -0.0174 19 Gadiformes 0.022 0.0181 0.0237 0.0325 0.0955 -0.0038 -0.013 0.0138 0.0979 -0.0229 0.0482 20 Sculpins/ Zoarcids 0.0068 0.0059 0.0061 0.0125 0.0574 -0.0012 -0.0089 -0.0153 -0.0002 0.0074 0.0268 21 Capelin 0.0566 0.0644 0.0603 0.0523 0.0412 -0.0131 0.0002 -0.0188 0.0609 0.0854 0.117 22 Sandlance 0.0601 0.0606 0.0555 0.0243 -0.0197 -0.0101 -0.0015 -0.0122 -0.0557 0.0973 0.215 23 Sharks/Rays 1E-05 2E-05 3E-06 0.0023 -0.0008 -0.0003 -5E-05 -0.0004 -0.0001 -0.0005 0.0001 24 Other Marine Fish 0.0147 0.0132 0.0146 0.0165 0.049 -0.0037 0.0149 0.0292 0.0147 0.062 0.0408 25 Brackish Fish 0.0002 -0.0002 0.0007 0.0024 0.0058 -0.0002 7E-06 -0.0008 0.0135 0.0497 -0.0069 26 Cephalopods -0.0129 -0.0122 -0.0135 -0.0046 0.0182 0.0024 0.0036 0.0016 -0.0191 -0.0116 -0.0382 27 MacroZoopl. 0.0044 0.0058 0.0042 0.0043 0.0235 -0.0113 0.0035 0.0005 0.0005 0.0151 0.0072 28 Euphausids 0.0178 0.0185 0.0169 0.0182 0.022 0.072 0.003 -0.0028 0.0022 0.0299 0.039 29 Copepods 0.0125 0.0133 0.0117 0.0118 -0.0064 0.196 -0.0045 -0.011 -0.0237 0.0114 0.0405 30 Crustaceans 0.0281 0.0282 0.0346 0.0269 0.0683 -0.0015 -0.0746 -0.0722 0.145 0.0441 0.0765 31 Other MesoZoopl. -0.0128 -0.0126 -0.0138 -0.0135 -0.0239 -0.0624 0.0227 0.0259 -0.0348 -0.0175 -0.0307 32 MicroZoopl. 0.0017 0.0029 0.0018 0.0009 -0.0014 -0.0023 0.0113 0.0094 -0.012 0.0129 0.0063 33 Marine Worms -0.0124 -0.0063 -0.0046 0.0036 0.0081 -0.0028 0.0247 0.0427 0.008 0.0173 -0.0092 34 Echinoderms 0.0012 0.006 0.0067 0.0032 0.0047 -0.0027 0.0779 0.0336 0.0618 0.0111 -0.0149 35 Bivalves -0.014 -0.007 -0.0081 0.0027 0.023 0.0072 0.154 0.213 0.0224 0.0208 -0.0548 36 Other Benthos 0.0354 0.013 0.0135 0.011 0.0121 0.0212 -0.0155 -0.028 0.0363 0.0176 0.0089 37Primary Production 0.0325 0.0351 0.0331 0.0274 0.0233 0.113 0.0062 -5E-05 0.0147 0.0649 0.0873 38 Ice Algae 0.0151 0.0121 0.0131 0.0149 0.0329 0.0335 0.0183 0.0236 0.0378 0.0253 0.0218 39 Ice Detritus 0.0109 0.0064 0.0082 0.0179 0.0474 0.0158 0.136 0.162 0.0882 0.0455 -0.032 40 Pelagic Detritus 0.0094 0.0106 0.01 0.0082 0.0117 0.0075 -0.0046 -0.0055 0.0129 0.0198 0.0248 41 SH Polar Bear 0.009 -0.352 0.0111 0.0125 -0.0037 -0.0018 -0.0005 0.069 0.0475 0.0119 0.0167 42 WHB Polar Bear -0.287 0.0145 0.02 0.0155 -0.0084 -0.0022 0.0081 -0.0058 0.0387 -0.0005 0.0371 43 FB Polar Bear 0.0099 0.0079 -0.347 0.0088 -0.0041 -0.0012 -0.0003 -0.0032 0.0377 0.131 0.0162 44 Killer whale -0.0047 -0.0052 -0.0036 -0.492 0.063 0.0726 0.0109 0.0937 0.0161 0.103 -0.026 45 Bowhead -6E-06 -1E-05 5E-06 -0.0052 0.0007 -0.425 0.0001 0.001 0.0002 0.0012 -0.0002 46 Narwhal 0.0005 0.0004 0.0005 -0.0162 -0.431 0.0024 0.0001 0.0034 0.0024 0.0041 0.0006 47 N Walrus -0.0011 0.0001 0.0002 -0.0066 0.0019 0.001 -0.473 0.0033 0.0016 0.0021 -0.0002 48 S Walrus 0.0259 0.0256 0.0237 0.0052 -0.0078 -0.0004 -0.0002 -0.308 -0.0224 -0.0407 0.0926 49 Beluga E 0.0003 -0.0009 0.0003 -0.0018 0.0014 0.0003 -0.0001 0.0005 0.001 0.0007 0.0005 50 Beluga W -0.0082 0.0028 -0.0057 -0.014 0.0078 0.0021 0.0002 0.0031 0.0099 0.0089 0.005 51 Beluga S 0.0002 -0.0049 0.0003 -0.0037 0.0006 0.0005 7E-05 0.0017 0.001 0.0012 0.0003 52 Sealing -0.0401 -0.0406 -0.0497 -0.0268 0.0112 0.0036 0.0013 0.012 -0.247 -0.135 -0.0788 53 Bird Hunting 3E-05 -0.0002 4E-05 3E-05 0.0002 -1E-05 7E-05 0.0001 0.0003 0.0004 0.0002 54 Fishing 0 0 0 0 0 0 0 0 0 0 0 From the Tropics to the Poles, Wabnitz and Hoover  APPENDIX 5 (CONTINUED): MIXED TROPHIC IMPACT RESULTS  Impacting / Impacted Harp seal Beluga  E Beluga  W Beluga James SeabirdsArctic Char Atlantic Salmon Gadi-formes Sculpins/ Zoarcids Capelin Sand- lance 1 Polar Bear WHB -0.113 0.0065 -0.15 0.0273 0.0038 0.0405 0.0142 0.0451 0.0226 0.021 0.0162 2 SH Polar Bear -0.111 -0.0252 0.0103 -0.284 0.0003 -0.0019 0.0018 0.0106 0.0079 0.009 0.0077 3 Polar Bear Foxe -0.0682 0.0027 -0.0474 0.012 0.0015 0.0133 0.005 0.0173 0.0096 0.0085 0.0072 4 Killer Whale -0.0261 -0.0089 -0.0381 -0.0383 0.0003 0.0106 0.0041 -0.0007 -0.0047 0.0008 -0.0105 5 Narwhal -0.0007 -0.002 -0.0038 -0.0022 -0.0008 -0.001 -0.0026 -0.0148 -0.013 8E-05 0.0052 6 Bowhead -0.0006 -0.0003 -0.0007 -0.0008 -0.0001 0.0002 -0.0001 4E-05 -4E-05 -0.0003 -0.0003 7 Walrus N -0.0001 -0.0004 -0.0015 -0.0002 -0.0004 0.0005 4E-05 -0.0045 -0.0034 0.001 0.001 8 Walrus S 0.089 0.0104 0.0409 0.0481 0.0041 -0.013 -0.0042 0.0664 0.0629 0.023 0.0755 9 Bearded Seal -0.0765 -0.003 -0.0274 -0.0278 0.0003 -0.0111 -0.0087 -0.0032 0.0063 0.0028 0.0143 10 Harbour Seal -0.0034 -0.0007 -0.0019 -0.0014 -0.0008 0.0007 0.0002 -0.0037 -0.005 -0.0029 -0.0045 11 Ringed Seal -0.291 -0.0363 -0.138 -0.164 -0.0142 0.0469 0.0156 -0.24 -0.217 -0.0757 -0.255 12 Harp seal -0.0252 -0.0013 -0.0084 -0.0135 -0.0008 0.0033 -0.003 0.0029 0.0016 -0.0145 0.0027 13 Beluga E -0.0013 -0.464 -0.001 -0.0017 -0.0004 0.0016 -0.0062 -0.0033 -0.0075 -0.0013 0.0022 14 Beluga W -0.0537 -0.0102 -0.271 -0.025 -0.0107 -0.191 -0.0609 -0.0933 -0.0091 -0.0491 0.0293 15 Beluga James -0.0164 -0.0025 -0.0026 -0.184 -0.0008 0.0019 -0.0019 -0.0011 0.0012 -0.0114 0.0014 16 Seabirds -0.0496 -0.0101 -0.0271 -0.0331 -0.521 -0.152 -0.166 0.0036 -0.0152 -0.0533 0.0004 17 Arctic Char -0.0123 -0.0036 0.0307 -0.0045 0.0083 -0.01 -0.0801 -0.0156 -0.019 -0.0118 -0.0112 18 Atlantic Salmon -0.0092 -0.0008 -0.0072 -0.0106 -0.0077 -0.0948 -0.112 -0.0163 -0.0314 -0.035 0.0024 19 Gadiformes -0.101 0.0284 0.0634 -0.0131 -0.0061 -0.0245 -0.0002 -0.141 -0.0629 -0.0903 -0.138 20 Sculpins/ Zoarcids -0.0664 0.0413 -0.0037 -0.0345 0.0044 -0.0049 0.0037 -0.0438 -0.114 -0.0689 -0.106 21 Capelin 0.71 0.0377 0.0971 0.333 0.0555 -0.0491 -0.025 -0.0247 -0.0115 -0.0906 -0.0592 22 Sandlance -0.108 -0.0132 -0.0388 -0.0639 0.0073 0.0301 0.0131 -0.0206 -0.0092 -0.0419 -0.0898 23 Sharks/Rays -0.0005 -4E-05 -0.0002 -0.0002 2E-06 5E-05 2E-05 1E-05 -1E-05 1E-05 -5E-05 24 Other Marine Fish -0.0116 -0.0061 0.0136 -0.033 0.0127 0.0073 0.0019 0.0364 0.0211 -0.0483 -0.0245 25 Brackish Fish -0.0142 0.0081 0.0245 -0.0084 0.0439 -0.0041 0.0047 -0.0011 0.0007 -0.0157 -0.006 26 Cephalopods -0.0497 0.0087 0.0102 0.0117 0.0305 -0.14 -0.101 -0.0988 -0.161 -0.0718 -0.0645 27 MacroZoopl. 0.0413 -0.015 0.0009 0.0035 0.0658 0.0123 0.0489 -0.0103 -0.0262 0.0551 -0.0313 28 Euphausids 0.0836 0.0657 0.065 0.101 0.0299 -0.0719 0.0614 -0.0038 -0.0012 0.12 0.039 29 Copepods 0.0325 0.0422 0.0403 0.0589 -0.0014 0.113 -0.0582 -0.0336 -0.0207 0.0579 0.134 30 Crustaceans 0.0502 0.0572 0.0395 0.0565 -0.0063 0.0417 0.138 -0.137 -0.0784 0.0517 0.0109 31 Other MesoZoopl. -0.0434 -0.0464 -0.0391 -0.0581 -0.001 0.0552 -0.0626 0.0283 0.0091 -0.0556 0.0095 32 MicroZoopl. 0.0155 -0.0108 -0.0019 -0.002 0.0196 0.0265 0.0872 -0.0123 -0.023 0.0246 0.0545 33 Marine Worms -0.0112 0.0158 0.0179 0.0084 -0.003 0.0059 -0.0119 0.065 0.0619 -0.0221 -0.018 34 Echinoderms -0.0204 -0.0149 -0.0116 -0.0245 0.0044 0.0143 -0.0122 0.0832 0.0449 -0.0129 -0.0118 35 Bivalves 0.0218 0.0018 0.0096 0.0042 0.039 -0.039 0.0017 0.0762 0.0648 0.0045 -0.004 36 Other Benthos -0.0418 0.0752 0.0566 0.0481 -0.0024 0.0019 -0.0118 0.0637 0.0701 -0.0286 -0.0164 37 Primary Production 0.142 0.0326 0.0497 0.0872 0.0463 0.156 0.114 -0.0558 -0.0499 0.2 0.218 38 Ice Algae -0.0064 0.0317 0.0336 0.0194 0.0165 0.0444 0.0437 0.0974 0.0602 -0.0013 0.01 39 Ice Detritus -0.0332 0.0654 0.0564 0.0314 0.0253 -0.0139 -0.0047 0.185 0.226 -0.0399 -0.0444 40 Pelagic Detritus 0.0394 0.0075 0.011 0.0201 0.0353 0.0036 0.0109 -0.0148 -0.0117 0.0522 0.035 41 SH Polar Bear 0.0816 0.0185 -0.0075 0.208 -0.0002 0.0014 -0.0013 -0.0077 -0.0058 -0.0066 -0.0056 42 WHB Polar Bear 0.069 -0.004 0.0918 -0.0167 -0.0023 -0.0247 -0.0086 -0.0275 -0.0138 -0.0128 -0.0099 43 FB Polar Bear 0.0487 -0.0019 0.0338 -0.0086 -0.0011 -0.0095 -0.0036 -0.0123 -0.0068 -0.0061 -0.0052 44 Killer whale 0.0261 0.0089 0.0381 0.0383 -0.0003 -0.0106 -0.0041 0.0007 0.0047 -0.0008 0.0105 45 Bowhead 0.0004 0.0002 0.0005 0.0006 8E-05 -0.0001 9E-05 -3E-05 3E-05 0.0002 0.0002 46 Narwhal 0.0005 0.0015 0.0029 0.0017 0.0006 0.0008 0.002 0.0114 0.0101 -6E-05 -0.004 47 N Walrus 9E-05 0.0004 0.0013 0.0002 0.0004 -0.0005 -3E-05 0.004 0.0031 -0.0009 -0.0009 48 S Walrus -0.0389 -0.0045 -0.0179 -0.021 -0.0018 0.0057 0.0018 -0.029 -0.0275 -0.01 -0.033 49 Beluga E 0.0011 -0.462 0.0008 0.0015 0.0003 -0.0014 0.0053 0.0028 0.0065 0.0011 -0.0019 50 Beluga W 0.0156 0.003 -0.212 0.0073 0.0031 0.0556 0.0177 0.0271 0.0027 0.0143 -0.0085 51 Beluga S 0.0032 0.0005 0.0005 -0.158 0.0002 -0.0004 0.0004 0.0002 -0.0002 0.0022 -0.0003 52 Sealing -0.0951 0.006 0.0284 0.0325 0.002 -0.0031 0.0013 0.0322 0.0266 0.0118 0.0286 53 Bird Hunting 0.0008 0.0002 0.0004 0.0005 -0.0079 0.0025 0.0027 -6E-05 0.0003 0.0009 -7E-06 54 Fishing 0 0 0 0 0 -6E-06 0 0 -1E-06 0 0  60  An Ecosystem Model of Hudson Bay, Canada    APPENDIX 5 (CONTINUED): MIXED TROPHIC IMPACT RESULTS  Impacting / Impacted Sharks /Rays Other Marine Fish Brackish Fish Cephalo-pods Macro Zoopl Euph-ausids CopepodsCrust-aceans Other MesoZooplMicro Zoopl Marine Worms 1 Polar Bear WHB 0.0348 0.0013 0.00764 0.0118 -0.00699 -0.000107 0.000067 -0.000571 0.000405 0.000927 -0.001462 SH Polar Bear 0.0169 0.0011 -0.000756 0.00204 -0.00191 -0.000517 0.000146 -0.000137 0.00008 0.000171 -0.0005143 Polar Bear Foxe 0.0152 0.0012 0.00358 0.0041 -0.00268 -0.000186 0.000058 -0.000221 0.000185 0.000336 -0.0006144 Killer Whale -0.442 -0.0006 0.0034 0.00242 -0.000434 0.000964 -0.000157 0.000329 -0.000016 0.000134 0.0002865 Narwhal -0.0176 -0.0017 -0.00206 -0.0094 0.000082 -0.000294 0.00006 -0.000287 -0.00032 -0.000063 0.0009386 Bowhead -0.0064 -7E-05 -0.000079 -0.00003 -0.00062 -0.00157 0.000278 -0.000333 0.00025 -0.000041 -0.0000767 Walrus N -0.0084 -0.0047 0.000275 0.000006 -0.000319 -0.000392 0.000153 -0.000691 0.0013 0.000028 0.00112 8 Walrus S 0.0338 0.0067 -0.00558 0.00747 -0.00606 -0.00418 0.000731 -0.00214 0.000625 0.000487 -0.003819 Bearded Seal -0.0331 0.0033 -0.000227 0.00245 -0.00063 -0.000185 0.000021 -0.00214 -0.000432 0.000101 0.00050510 Harbour Seal -0.0078 -0.0034 -0.00641 -0.000618 0.00084 0.000261 -0.000056 0.000193 -0.00006 -0.000079 0.00019211 Ringed Seal -0.143 -0.0305 0.0195 -0.0259 0.02 0.0136 -0.00225 0.00645 -0.000522 -0.00164 0.0139 12 Harp seal -0.0007 -0.0013 0.00205 0.00102 0.00237 0.000436 -0.000271 0.000759 -0.0002 -0.000159 -0.00019713 Beluga E -0.0034 0.0026 -0.00164 -0.0033 0.000822 -0.000635 0.000132 -0.000683 -0.000166 -0.000165 -0.00006814 Beluga W -0.0706 0.008 -0.0408 -0.0419 0.0203 -0.0055 0.000837 0.000288 -0.00132 -0.00324 0.00018115 Beluga James -0.0175 0.0009 0.00111 -0.00285 0.00219 -0.000008 -0.000103 0.000073 -0.000149 -0.000226 -0.00025116 Seabirds -0.0299 -0.0327 -0.236 -0.114 0.0152 -0.0017 -0.00008 0.0183 0.00509 -0.00135 -0.00065917 Arctic Char 0.0341 -0.0256 -0.0647 0.0129 -0.0183 0.00363 -0.00232 -0.00415 -0.00454 0.00489 -0.0015 18 Atlantic Salmon -0.0081 -0.0387 -0.154 -0.128 -0.0393 0.00315 0.00268 -0.0253 0.00336 0.00216 0.00671 19 Gadiformes 0.0508 -0.256 -0.00137 0.0455 0.0292 0.00605 -0.000659 0.0111 0.00944 -0.00283 -0.0238 20 Sculpins/ Zoarcids 0.0882 -0.131 -0.00446 0.0431 0.0189 0.00561 -0.000275 -0.00769 0.00824 -0.00157 -0.0202 21 Capelin -0.0161 -0.0219 -0.0474 0.0177 -0.146 -0.0277 0.0155 -0.0388 0.0101 0.0109 0.00666 22 Sandlance 0.035 -0.0399 0.000597 -0.0891 -0.0386 -0.0346 0.00327 -0.0133 -0.0172 0.00279 0.00344 23 Sharks/Rays -0.0495 0 0.000016 0.000019 -0.000003 0.000004 -0.000001 0.000002 0 0.000001 0 24 Other Marine Fish 0.0417 -0.0374 -0.0205 -0.0359 -0.0149 0.00507 -0.00259 -0.0509 0.00377 0.00434 -0.0067925 Brackish Fish 0.0295 -0.0072 -0.0487 -0.0357 -0.0193 0.0039 0.000597 -0.00699 0.00122 0.00154 0.00037826 Cephalopods -0.0059 0.0435 0.0158 -0.109 -0.06 0.0206 -0.00206 -0.00334 -0.00398 0.00996 0.00791 27 MacroZoopl. 0.045 0.0223 0.0648 0.0915 -0.0561 -0.197 0.00774 -0.0666 -0.12 -0.143 0.0172 28 Euphausids -0.0014 -0.0224 0.0445 -0.019 -0.0601 -0.108 -0.322 0.00371 -0.0605 0.197 0.0104 29 Copepods -0.0102 0.0303 -0.116 -0.0116 -0.147 0.332 -0.375 -0.0674 -0.168 -0.46 0.0016 30 Crustaceans -0.0327 0.177 0.0907 0.0509 0.0256 -0.0682 -0.0194 -0.0531 0.0397 -0.00509 -0.166 31 Other MesoZoopl. 0.0131 -0.0423 -0.0191 0.0513 0.0154 -0.37 0.00929 -0.224 -0.0614 -0.151 0.0389 32 MicroZoopl. 0.0183 -0.0271 0.126 0.0719 0.16 -0.0289 -0.0932 -0.091 -0.0557 -0.137 0.0204 33 Marine Worms 0.0417 0.0041 -0.0126 -0.00539 -0.0413 -0.0104 0.000428 -0.0321 -0.0303 0.0019 -0.0851 34 Echinoderms 0.0944 -0.0644 0.00627 0.011 0.0338 -0.0276 0.009 -0.0813 0.0519 -0.00984 -0.171 35 Bivalves 0.0125 0.009 -0.0238 -0.0118 0.0267 0.0689 -0.0202 0.0605 -0.205 0.00149 -0.149 36 Other Benthos 0.0049 -0.011 0.0296 -0.0114 -0.149 0.0276 -0.00383 0.0606 0.0298 0.0173 -0.125 37 Primary Production 0.0265 0.119 0.116 0.0789 0.191 0.111 0.321 0.0412 0.231 0.257 0.00847 38 Ice Algae 0.0332 0.0772 0.0237 0.0335 0.069 0.0318 0.0802 0.0762 0.137 0.0249 0.0343 39 Ice Detritus 0.0836 -0.0305 0.00629 -0.00335 -0.0852 0.0428 -0.0147 0.124 -0.116 0.00883 0.282 40 Pelagic Detritus 0.0037 0.0783 0.0494 0.00488 0.0439 0.0128 0.0105 0.0643 0.0302 0.0396 -0.0117 41 SH Polar Bear -0.0124 -0.0008 0.000555 -0.0015 0.0014 0.000379 -0.000107 0.000101 -0.000058 -0.000125 0.00037742 WHB Polar Bear -0.0212 -0.0008 -0.00466 -0.00723 0.00426 0.000065 -0.000041 0.000349 -0.000247 -0.000566 0.00089 43 FB Polar Bear -0.0109 -0.0008 -0.00256 -0.00293 0.00192 0.000133 -0.000042 0.000158 -0.000133 -0.00024 0.00043944 Killer whale 0.442 0.0006 -0.0034 -0.00242 0.000434 -0.000964 0.000157 -0.000329 0.000016 -0.000134 -0.00028645 Bowhead 0.0047 5E-05 0.000059 0.000022 0.00046 0.00117 -0.000206 0.000247 -0.000186 0.00003 0.00005646 Narwhal 0.0136 0.0013 0.00159 0.00725 -0.000064 0.000227 -0.000047 0.000222 0.000247 0.000049 -0.00072447 N Walrus 0.0076 0.0043 -0.000248 -0.000005 0.000288 0.000354 -0.000138 0.000624 -0.00117 -0.000025 -0.0010148 S Walrus -0.0148 -0.0029 0.00244 -0.00326 0.00265 0.00183 -0.000319 0.000934 -0.000273 -0.000213 0.00166 49 Beluga E 0.0029 -0.0023 0.00142 0.00285 -0.000709 0.000548 -0.000114 0.00059 0.000143 0.000142 0.00005850 Beluga W 0.0205 -0.0023 0.0119 0.0122 -0.00589 0.0016 -0.000243 -0.000084 0.000384 0.000942 -0.00005351 Beluga S 0.0034 -0.0002 -0.000216 0.000553 -0.000425 0.000001 0.00002 -0.000014 0.000029 0.000044 0.00004952 Sealing 0.0308 0.0038 -0.00143 0.0025 -0.00292 -0.00181 0.000338 -0.000297 0.000252 0.000219 -0.0019653 Bird Hunting 0.0005 0.0005 0.00389 0.00187 -0.00025 0.000028 0.000001 -0.000302 -0.000084 0.000022 0.00001154 Fishing 0 0 0 0 0 0 0 0 0 0 0  From the Tropics to the Poles, Wabnitz and Hoover  APPENDIX 5 (CONTINUED): MIXED TROPHIC IMPACT RESULTS  *H indicates harvest of a species or group  Impacting / Impacted Marine Worms Echino -derms Bivalves Other Benthos Primary Production Ice Algae Ice Detritus H: SH  Polar Bear H: WHB  Polar BearH: FB  Polar Bear1 Polar Bear WHB -0.00146 -0.00474 -0.00213 0.000225 -0.000194 0.000011 0.00115 -0.0237 0.47 -0.0328 2 SH Polar Bear -0.000514 -0.000941 -0.000333 -0.000068 -0.000105 -0.00003 0.000275 0.48 -0.0123 -0.0152 3 Polar Bear Foxe -0.000614 -0.00182 -0.000929 0.00007 -0.000089 -0.000002 0.000497 -0.0111 -0.0138 0.485 4 Killer Whale 0.000286 0.000445 0.000462 0.000048 0.000048 0.000002 -0.000265 0.00521 0.0047 0.00359 5 Narwhal 0.000938 0.00203 0.000961 0.000139 -0.000001 -0.000028 -0.000635 -0.000523 -0.000615 -0.0007076 Bowhead -0.000076 0.000026 0.00007 -0.000092 -0.000133 -0.000087 0.000033 0.000015 0.000008 -0.0000077 Walrus N 0.00112 -0.00566 -0.00505 0.000927 -0.000138 -0.000009 0.00157 -0.000164 0.00118 -0.000258 Walrus S -0.00381 -0.00912 -0.00588 0.000031 -0.00042 0.000076 0.00322 -0.0585 -0.0593 -0.0543 9 Bearded Seal 0.000505 -0.000473 0.000655 0.000257 0.000042 0.000064 -0.000333 0.042 0.0417 0.0752 10 Harbour Seal 0.000192 0.000272 0.000358 -0.000005 0.00004 0.000005 -0.000173 -0.000704 -0.000728 0.00371 11 Ringed Seal 0.0139 0.0286 0.0137 0.00111 0.00124 -0.000289 -0.00912 0.197 0.194 0.172 12 Harp seal -0.000197 -0.000625 -0.000255 -0.000058 0.00017 0.000107 0.000149 0.0112 0.0113 0.0157 13 Beluga E -0.000068 0.000994 0.000532 -0.000105 -0.000016 -0.000008 -0.000149 0.001 -0.000331 -0.00035914 Beluga W 0.000181 0.00979 0.00388 -0.00165 0.000181 -0.000035 -0.00124 -0.0095 0.0282 0.0194 15 Beluga James -0.000251 0.000057 0.000058 -0.000099 0.000104 0.000062 0.000058 0.0255 -0.00128 -0.0015316 Seabirds -0.000659 -0.0114 -0.0135 -0.00209 -0.000438 -0.000382 0.00562 0.0102 -0.00194 -0.0025317 Arctic Char -0.0015 0.000075 0.00347 0.000516 0.000817 0.00119 -0.00094 -0.00239 -0.000871 -0.00043318 Atlantic Salmon 0.00671 0.0146 0.00669 0.00328 -0.00136 -0.00113 -0.00464 -0.00597 -0.00578 -0.0056619 Gadiformes -0.0238 -0.0756 -0.0248 -0.00416 0.000182 0.000152 0.0178 0.0181 0.022 0.0237 20 Sculpins/Zoarcids -0.0202 -0.0349 -0.0216 -0.00201 0.000212 0.000765 0.0131 0.0059 0.00682 0.00614 21 Capelin 0.00666 0.0192 0.00843 0.00289 -0.01 -0.00575 -0.00487 0.0644 0.0566 0.0603 22 Sandlance 0.00344 0.00754 0.00346 0.00203 -0.000642 0.00147 -0.0025 0.0606 0.0601 0.0555 23 Sharks/Rays 0 0 0.000001 0 0 0 -0.000001 0.000015 0.000012 0.00000324 Other Marine Fish -0.00679 0.0191 -0.00496 0.00804 0.0014 0.00104 0.0011 0.0132 0.0147 0.0146 25 Brackish Fish 0.000378 0.000828 0.00153 -0.000634 -0.000453 -0.000078 -0.000188 -0.0002 0.000173 0.00069126 Cephalopods 0.00791 0.0183 0.00747 0.000939 -0.000135 0.000357 -0.00501 -0.0122 -0.0129 -0.0135 27 MacroZoopl. 0.0172 0.0235 0.00188 0.0192 0.0287 0.0151 -0.00961 0.00581 0.00442 0.00418 28 Euphausids 0.0104 0.0149 0.00241 0.0176 0.136 0.112 -0.0101 0.0185 0.0178 0.0169 29 Copepods 0.0016 0.0178 -0.00916 -0.0331 -0.248 -0.213 0.0163 0.0133 0.0125 0.0117 30 Crustaceans -0.166 -0.393 -0.166 -0.127 -0.0114 -0.007 0.123 0.0282 0.0281 0.0346 31 Other MesoZoopl. 0.0389 0.089 0.0684 0.02 -0.0257 -0.0646 -0.0357 -0.0126 -0.0128 -0.0138 32 MicroZoopl. 0.0204 0.0534 0.0353 -0.00803 -0.157 -0.0898 -0.0131 0.00286 0.00173 0.00175 33 Marine Worms -0.0851 -0.0163 -0.00349 -0.305 0.00233 0.00545 -0.07 -0.00625 -0.0124 -0.0045934 Echinoderms -0.171 0.0119 -0.193 -0.042 -0.0036 0.00449 0.0713 0.006 0.00123 0.0067 35 Bivalves -0.149 -0.052 -0.179 -0.109 0.0198 0.00427 -0.226 -0.00703 -0.014 -0.0080936 Other Benthos -0.125 -0.148 -0.236 -0.143 -0.00725 -0.0593 -0.191 0.013 0.0354 0.0135 37 Primary Production 0.00847 0.00831 0.027 -0.0259 -0.28 -0.23 -0.0039 0.0351 0.0325 0.0331 38 Ice Algae 0.0343 0.00388 0.019 0.115 -0.069 -0.074 -0.0575 0.0121 0.0151 0.0131 39 Ice Detritus 0.282 0.225 0.318 0.273 0.00791 -0.0316 0 0.00642 0.0109 0.00821 40 Pelagic Detritus -0.0117 -0.026 -0.0111 -0.0117 -0.0209 -0.0167 0.00944 0.0106 0.00944 0.00999 41 SH Polar Bear 0.000377 0.00069 0.000244 0.00005 0.000077 0.000022 -0.000202 -0.352 0.00904 0.0111 42 WHB Polar Bear 0.00089 0.00289 0.0013 -0.000137 0.000118 -0.000007 -0.000703 0.0145 -0.287 0.02 43 FB Polar Bear 0.000439 0.0013 0.000664 -0.00005 0.000064 0.000001 -0.000355 0.00793 0.00989 -0.347 44 Killer whale -0.000286 -0.000445 -0.000462 -0.000048 -0.000048 -0.000002 0.000265 -0.00521 -0.0047 -0.0035945 Bowhead 0.000056 -0.000019 -0.000052 0.000068 0.000098 0.000065 -0.000025 -0.000011 -0.000006 0.00000546 Narwhal -0.000724 -0.00157 -0.000742 -0.000107 0.000001 0.000022 0.00049 0.000404 0.000475 0.00054647 N Walrus -0.00101 0.00512 0.00456 -0.000838 0.000125 0.000009 -0.00142 0.000148 -0.00106 0.00022648 S Walrus 0.00166 0.00398 0.00257 -0.000014 0.000183 -0.000033 -0.0014 0.0256 0.0259 0.0237 49 Beluga E 0.000058 -0.000857 -0.000459 0.000091 0.000014 0.000007 0.000128 -0.000867 0.000285 0.00030950 Beluga W -0.000053 -0.00284 -0.00113 0.000478 -0.000053 0.00001 0.00036 0.00276 -0.0082 -0.0056551 Beluga S 0.000049 -0.000011 -0.000011 0.000019 -0.00002 -0.000012 -0.000011 -0.00494 0.000248 0.00029852 Sealing -0.00196 -0.00347 -0.002 -0.000215 -0.000209 -0.000002 0.00129 -0.0406 -0.0401 -0.0497 53 Bird Hunting 0.000011 0.000188 0.000221 0.000034 0.000007 0.000006 -0.000092 -0.000168 0.000032 0.00004254 Fishing 0 0 0 0 0 0 0 0 0 0 62  An Ecosystem Model of Hudson Bay, Canada    APPENDIX 5 (CONTINUED): MIXED TROPHIC IMPACT RESULTS *H indicates harvest of a species or group  Impacting / Impacted H : Killer Whale H:  Bowhead H:  Narwhal H: N Walrus H: S WalrusH: Beluga W H: Beluga S H: Sealing H: Bird  Hunting H: Fishing1 Polar Bear WHB -0.0255 0.00353 0.0138 -0.0133 0.00945 -0.15 0.0273 -0.0626 0.00376 0.0319 2 SH Polar Bear -0.017 0.0024 0.00503 0.000669 -0.094 0.0103 -0.284 -0.0371 0.000305 0.00439 3 Polar Bear Foxe -0.0123 0.0017 0.00572 0.000446 0.00449 -0.0474 0.012 -0.033 0.00148 0.0117 4 Killer Whale 0.492 -0.0726 -0.063 -0.0109 -0.0937 -0.0381 -0.0383 0.0121 0.000251 0.00259 5 Narwhal 0.021 -0.00312 0.558 -0.000182 -0.00434 -0.00378 -0.00221 -0.00144 -0.000815 -0.00615 6 Bowhead 0.007 0.573 -0.000999 -0.000159 -0.00133 -0.000719 -0.000799 0.000061 -0.000109 0.000025 7 Walrus N 0.00733 -0.00115 -0.00212 0.523 -0.00368 -0.00145 -0.000176 -0.000349 -0.000407 -0.00156 8 Walrus S -0.0119 0.000889 0.0179 0.000386 0.706 0.0409 0.0481 -0.127 0.00405 0.0271 9 Bearded Seal 0.0464 -0.00696 -0.00565 -0.00211 -0.0177 -0.0274 -0.0278 0.231 0.00034 -0.00259 10 Harbour Seal 0.00629 -0.000867 -0.00219 -0.000111 -0.00134 -0.00193 -0.00139 0.000641 -0.000755 -0.00232 11 Ringed Seal 0.067 -0.00701 -0.0671 -0.00352 -0.041 -0.138 -0.164 0.395 -0.0142 -0.095 12 Harp seal 0.0114 -0.0016 -0.00175 -0.000682 -0.00461 -0.00837 -0.0135 0.0151 -0.000761 0.000638 13 Beluga E 0.00213 -0.000386 -0.00159 0.00014 -0.000562 -0.000965 -0.00173 -0.000778 -0.000393 -0.00168 14 Beluga W 0.0481 -0.00732 -0.0267 -0.000766 -0.0107 0.729 -0.025 -0.0231 -0.0107 -0.0964 15 Beluga James 0.0189 -0.00277 -0.00325 -0.000371 -0.0088 -0.00262 0.816 -0.00317 -0.000805 -0.00038 16 Seabirds -0.00177 0.000689 -0.0149 -0.00439 -0.00732 -0.0271 -0.0331 -0.0136 0.479 -0.0718 17 Arctic Char 0.000833 -0.00061 -0.00183 -0.000096 -0.00116 0.0307 -0.00452 -0.00251 0.00826 0.349 18 Atlantic Salmon -0.00393 0.000338 -0.0167 0.00241 0.00186 -0.00724 -0.0106 -0.0122 -0.00774 -0.044 19 Gadiformes 0.0325 -0.00378 0.0955 -0.013 0.0138 0.0634 -0.0131 0.0587 -0.00607 0.13 20 Sculpins/Zoarcids 0.0125 -0.00124 0.0574 -0.00893 -0.0153 -0.00373 -0.0345 0.0167 0.00436 0.156 21 Capelin 0.0523 -0.0131 0.0412 0.000176 -0.0188 0.0971 0.333 0.115 0.0555 0.0651 22 Sandlance 0.0243 -0.0101 -0.0197 -0.00152 -0.0122 -0.0388 -0.0639 0.128 0.00725 0.0269 23 Sharks/Rays 0.00232 -0.000342 -0.000778 -0.000051 -0.000439 -0.000171 -0.00017 0.00003 0.000002 0.000017 24 Other Marine Fish 0.0165 -0.00374 0.049 0.0149 0.0292 0.0136 -0.033 0.032 0.0127 0.0586 25 Brackish Fish 0.00242 -0.000235 0.00578 0.000007 -0.00075 0.0245 -0.00841 -0.000973 0.0439 0.0161 26 Cephalopods -0.00463 0.00242 0.0182 0.00362 0.00163 0.0102 0.0117 -0.0329 0.0305 -0.113 27 MacroZoopl. 0.00427 -0.0113 0.0235 0.00348 0.000501 0.000905 0.00347 0.0061 0.0658 0.00467 28 Euphausids 0.0182 0.072 0.022 0.00303 -0.00278 0.065 0.101 0.0294 0.0299 -0.0131 29 Copepods 0.0118 0.196 -0.00637 -0.00448 -0.011 0.0403 0.0589 0.0216 -0.00135 0.0388 30 Crustaceans 0.0269 -0.00153 0.0683 -0.0746 -0.0722 0.0395 0.0565 0.0955 -0.00634 -0.0114 31 Other MesoZoopl. -0.0135 -0.0624 -0.0239 0.0227 0.0259 -0.0391 -0.0581 -0.0322 -0.001 0.0192 32 MicroZoopl. 0.000926 -0.00227 -0.00139 0.0113 0.00944 -0.00188 -0.00196 0.00122 0.0196 0.0087 33 Marine Worms 0.00361 -0.0028 0.00807 0.0247 0.0427 0.0179 0.00843 -0.00416 -0.00297 0.0255 34 Echinoderms 0.00324 -0.00272 0.00472 0.0779 0.0336 -0.0116 -0.0245 0.00726 0.00441 0.0268 35 Bivalves 0.00269 0.0072 0.023 0.154 0.213 0.00961 0.00415 -0.0303 0.039 0.0155 36 Other Benthos 0.011 0.0212 0.0121 -0.0155 -0.028 0.0566 0.0481 0.0156 -0.00242 0.025 37 Primary Production 0.0274 0.113 0.0233 0.00623 -0.000046 0.0497 0.0872 0.0675 0.0463 0.0724 38 Ice Algae 0.0149 0.0335 0.0329 0.0183 0.0236 0.0336 0.0194 0.0258 0.0165 0.0541 39 Ice Detritus 0.0179 0.0158 0.0474 0.136 0.162 0.0564 0.0314 0.00311 0.0253 0.0736 40 Pelagic Detritus 0.00815 0.00745 0.0117 -0.00455 -0.00547 0.011 0.0201 0.0217 0.0353 0.00754 41 H: SH Polar Bear 0.0125 -0.00176 -0.00369 -0.000491 0.069 -0.00754 0.208 0.0272 -0.000224 -0.00322 42 H: WHB Polar Bear 0.0155 -0.00215 -0.00844 0.00814 -0.00577 0.0918 -0.0167 0.0382 -0.00229 -0.0195 43 H:FB Polar Bear 0.00876 -0.00122 -0.00409 -0.000319 -0.00321 0.0338 -0.00857 0.0236 -0.00106 -0.00839 44 H: Killer whale -0.492 0.0726 0.063 0.0109 0.0937 0.0381 0.0383 -0.0121 -0.000251 -0.00259 45 H : Bowhead -0.0052 -0.425 0.000741 0.000118 0.000989 0.000534 0.000593 -0.000045 0.000081 -0.00001846 H : Narwhal -0.0162 0.00241 -0.431 0.00014 0.00335 0.00292 0.0017 0.00111 0.000629 0.00475 47 H : N Walrus -0.00662 0.00104 0.00192 -0.473 0.00332 0.00131 0.000159 0.000315 0.000368 0.00141 48 H : S Walrus 0.00522 -0.000388 -0.00784 -0.000169 -0.308 -0.0179 -0.021 0.0556 -0.00177 -0.0119 49 H : Beluga E -0.00183 0.000333 0.00137 -0.00012 0.000485 0.000832 0.00149 0.000672 0.000339 0.00145 50 H : Beluga W -0.014 0.00213 0.00776 0.000223 0.0031 -0.212 0.00727 0.00671 0.00311 0.028 51 H : Beluga S -0.00366 0.000537 0.000632 0.000072 0.00171 0.000509 -0.158 0.000615 0.000156 0.000074 52 H : Sealing -0.0268 0.00359 0.0112 0.00127 0.012 0.0284 0.0325 -0.128 0.00199 0.0134 53 H : Bird Hunting 0.000029 -0.000011 0.000245 0.000072 0.00012 0.000445 0.000545 0.000223 -0.00787 0.00118 54 H : Fishing 0 0 0 0 0 0 0 0 0 -0.000003From the Tropics to the Poles, Wabnitz and Hoover   APPENDIX 6: COEFFICIENT OF VARIATION (CV) VALUES USED FOR MONTE CARLO ROUTINE  Functional Group Biomass (CV) P/B (CV) EE (CV) BA (CV) 1 Polar Bear WHB 0.15 0.25 0.1 0.05 2 SH Polar Bear 0.15 0.25 0.1 0.05 3 Polar Bear Foxe 0.15 0.25 0.1 0.05 4 Killer Whale 0.15 0.1 0.1 0.05 5 Narwhal 0.15 0.1 0.1 0.05 6 Bowhead 0.4 0.1 0.1 0.15 7 Walrus N 0.25 0.1 0.1 0.05 8 Walrus S 0.25 0.1 0.1 0.05 9 Bearded Seal 0.25 0.1 0.1 0.05 10 Harbour Seal 0.25 0.1 0.1 0.05 11 Ringed Seal 0.25 0.1 0.1 0.05 12 Harp seal 0.25 0.1 0.1 0.05 13 Beluga E 0.15 0.1 0.1 0.15 14 Beluga W 0.15 0.1 0.1 0.15 15 Beluga James 0.15 0.1 0.1 0.05 16 Seabirds 0.4 0.3 0.1 0.05 17 Arctic Char 0.1 0.2 0.1 0.05 18 Atlantic Salmon 0.1 0.2 0.1 0.05 19 Gadiformes 0.1 0.2 0.1 0.05 20 Sculpins/ Zoarcids 0.1 0.2 0.1 0.05 21 Capelin 0.1 0.2 0.1 0.05 22 Sandlance 0.1 0.2 0.1 0.05 23 Sharks/Rays 0.1 0.2 0.1 0.05 24 Other Marine Fish 0.1 0.2 0.1 0.05 25 Brackish Fish 0.1 0.25 0.1 0.05 26 Cephalopods 0.25 0.3 0.1 0.05 27 MacroZooplankton 0.25 0.2 0.1 0.05 28 Euphausids 0.15 0.2 0.1 0.05 29 Copepods 0.15 0.2 0.1 0.05 30 Crustaceans 0.15 0.2 0.1 0.05 31 Other MesoZooplankton 0.15 0.2 0.1 0.05 32 MicroZooplankton 0.25 0.35 0.1 0.05 33 Marine Worms 0.1 0.25 0.1 0.05 34 Echinoderms 0.1 0.25 0.1 0.05 35 Bivalves 0.1 0.25 0.1 0.05 36 Other Benthos 0.1 0.25 0.1 0.05 37 Primary Production 0.15 0.1 0.1 0.05 38 Ice Algae 0.15 0.1 0.1 0.05 64  An Ecosystem Model of Hudson Bay, Canada    APPENDIX 7: MONTE CARLO RESULTS FOR ESTIMATES OF BIOMASS (T·KM-2) AND P/B (YEAR-1)          From the Tropics to the Poles, Wabnitz and Hoover  APPENDIX 7 (CONTINUED): MONTE CARLO RESULTS FOR ESTIMATES OF BIOMASS (T·KM-2) AND P/B (YEAR-1)          66  An Ecosystem Model of Hudson Bay, Canada    APPENDIX 7 (CONTINUED): MONTE CARLO RESULTS FOR ESTIMATES OF BIOMASS (T·KM-2) AND P/B (YEAR-1)      From the Tropics to the Poles, Wabnitz and Hoover  APPENDIX 8: ECOSIM OUTPUT BIOMASS TRENDS FROM 1970-2010 AS SCALED TO 1970 BIOMASS VALUE.  68  An Ecosystem Model of Hudson Bay, Canada    APPENDIX 8 (CONTINUED)  ECOSIM OUTPUT BIOMASS TRENDS FROM 1970-2010 AS SCALED TO 1970 BIOMASS VALUE.  From the Tropics to the Poles, Wabnitz and Hoover  BASELINE TROPHIC RELATIONSHIPS IN KALOKO-HONOKŌHAU, HAWAI‘I 7 Colette C. C. Wabnitz Secretariat of the Pacific Community, BP D.5, 98848 Nouméa cedex, New Caledonia Email: Colette.Wabnitz@gmail.com ABSTRACT The formal protection of the Hawaiian green turtle (Chelonia mydas) in the 1970s has led to significant increases in the number of individuals recorded throughout the archipelago. However, reduced growth rates and poor body condition of individuals at a number of foraging sites, including Kaloko-Honokōhau National Historical Park (Kaloko, Big Island), suggest that some aggregations have reached carrying capacity. To better understand the ecological structure and processes of the reef system at the park, an ecosystem model was developed that synthesized available data on Kaloko for the year 2005 and included 26 groups, spanning the entire trophic web. Model results showed that the combined grazing pressure by the different herbivorous functional groups (i.e., reef fish, sea urchins and green turtles) matched overall algal production. Sea urchins exerted the strongest control over algal resources, partly because of their large biomass within park waters. Results confirmed that the Kaloko green turtle aggregation has reached carrying capacity. Green turtles help maintain low algal cover, and thus resilience of reefs in the face of disturbance and should be explicitly included in studies of ecosystem dynamics on reefs. The model also serves as a ‘current-condition’ baseline for Kaloko, and provides a valuable tool for the assessment of the future marine ecosystem impacts of projected urban expansion plans around the park.  INTRODUCTION Grazing by macroherbivores is one of the major processes structuring benthic coral reef communities. Studies which have highlighted the role of herbivores in promoting reef resilience and recovery to coral-dominated states where disturbance has led to increased algal biomass (e.g., Bellwood et al. 2004 ), have focused almost exclusively on fish and sea urchins (Hay 1984a; Hay1984b; Carreiro-Silva and McClanahan 2001; Mumby et al. 2006a; Paddack et al. 2006; Albert et al. 2008). Numerous reefs in the Caribbean have transitioned from coral to algal-dominated states (Gardner et al. 2003). This shift has been attributed to either a dramatic reduction in fish stocks, shown to limit the distribution, abundance and production of algae (Ogden and Lobel 1978; Hay 1981; Lewis 1986) or the region-wide loss to disease of an important echinoid herbivore (Diadema antillarum) in systems that had seen a shift from fish-dominated to echinoid-dominated herbivory (Lessios 1988; Mumby et al. 2006b). At some locations, increased anthropogenic nutrient loading has further exacerbated the abundance of primary producers (Burkepile and Hay 2006; Littler and Littler 2007). The majority of reefs in Hawai‘i are not as severely impacted as those throughout the Caribbean region (Brainard et al. 2002; Waddell and Clarke 2008). However, a number of locations are showing increasing signs of stress as a result of mounting anthropogenic pressures on the coastal zone through development and runoff, tourism and recreation activities, and overfishing (Grigg 1994; Hunter and Evans 1995; Friedlander et al. 2008; Williams et al. 2008). Few studies have focused on the role of sea turtles in maintaining coral reef resilience. In the context of the Caribbean this is understandable given that green turtles (Chelonia mydas) in this region mostly forage on seagrass (Bjorndal 1980; Mortimer 1981; Thayer et al. 1984) and thus have predominantly an indirect impact on the trophodynamics of reef systems (Valentine et al. 2002; Heck and Valentine 2006). Moreover, populations of the green turtle have been subject to a long history of human exploitation for eggs, turtle meat and shells (Parsons 1962), and in some Caribbean locations still suffer high harvest rates  (Campbell and Lagueux 2005). As a consequence their numbers have been so dramatically reduced that they probably no longer perform their functional role as grazers within local seagrass systems (Jackson                                                  7 Cite as: Wabnitz,C.C.C. 2012. Baseline Trophic Relationships in Kolako Honokōhau, Hawai‛i, p.69-107. In: Wabnitz, C.C.C. and Hoover, C. (eds.) From the Tropics to the Poles: Ecosystem Models of Hudson Bay, Kaloko- Honokōhau, Hawai‛i, and the Antarctic Peninsula Fisheries Centre Research Reports 20(2). Fisheries Centre, University of British Columbia [ISSN 1198-6727]. 70                  Baseline Trophic Relationships in Kaloko-Honokōkau, Hawai’i   1997; Bjorndal and Bolten 2003; Bjorndal and Jackson 2003). In Hawai‘i, however, green turtles feed primarily on algal species that commonly occur on the reef (McCutcheon et al. 2003; McDermid et al. 2007; Arthur and Balazs 2008; Russell and Balazs 2009) and may therefore play a direct role in maintaining the resilience of coral ecosystems in this region. Observations of green turtles actively feeding on Acanthophora spp. and Hypnea spp., both non-native algae on a number of Hawaiian reefs (Russell and Balazs 1994; Arthur and Balazs 2008; Russell and Balazs 2009) further highlight their contribution to the promotion of reef resilience.  Since turtle harvesting ended in the late 1970s (Witzell 1994), an approximately linear increase in abundance of nesting females has been observed at French Frigate Shoals, Northwestern Hawaiian Islands (Balazs and Chaloupka 2004a; Balazs and Chaloupka 2006), which accounts for > 90% of all nesting within the Hawaiian Archipelago (Balazs 1980). This increase in abundance is interpreted as a recovery trend because the Hawaiian green nesting population is one of several sea turtle stocks that has been continuously monitored using dependable methodology for several decades (Balazs and Chaloupka 2004a; Chaloupka et al. 2008) and is considered ‘self-contained’ (Dutton et al. 2008). Population trends at a number of foraging grounds that have been also subject to long term monitoring seem to mirror this trajectory (Chaloupka and Balazs 2007). The significant increase in green turtle abundance within the archipelago over the last 10–20 years has been associated with a significant decrease in somatic growth rates suggesting that the carrying capacity for a number of foraging grounds may have been reached (Balazs and Chaloupka 2004b).  The role of green turtles in maintaining algal communities in a low biomass/cropped state, at least in Hawai‘i, may be equally important to that of other grazers. An understanding of the ecological role of green turtles as grazers on reefs requires a process-oriented approach that assesses the relative contributions of all herbivorous functional groups (i.e., fish, urchins and green turtles). Such an approach is currently lacking, though it may provide significant insights into the need for, and consequences of, improved turtle conservation and management. Ecological modeling has developed ways to mathematically describe the complexity and non-linear behavior of ecological systems. Ecopath with Ecosim is a freely available, widely used software for describing the structure of ecosystems and their food webs. It was recently named as one of the 10 major scientific breakthroughs in the 200 year history of the US National Oceanographic and Atmospheric Administration (NOAA; see http://celebrating200years.noaa.gov/breakthroughs/welcome.html). Rather than providing outputs at the population level of biological organization, typical of many models, the Ecopath with Ecosim approach provides outputs at the ecosystem level, reflecting food-web linkages, energy cycling, and changes in biomass of each species group defined in the model (Christensen 2008). Although determining carrying capacity of a system has been highlighted as one of the uses of this software, few studies have explored this aspect (but see Christensen and Pauly 1998).  An Ecopath trophic model was developed to investigate the role that green turtles play in the coral reef ecosystem of Kaloko-Honokōhau National Historical Park (Kaloko). Located on the west coast of Hawai‘i Island (the ‘Big Island’; figure 1), the park supports a healthy and relatively diverse coral habitat with little evidence of non-native species of macroalgae or diseased coral (Gibbs et al. 2007). It has low fish biomass, but high fish diversity (Parrish et al. 1990, Beets et al. 2010). The park also has a resident foraging population of immature green turtles that has been the subject of a mark–recapture study by the NOAA National Marine Fisheries Service and the US National Park Service since 1999 (G. Balazs, Pacific Islands Fisheries Science Center, and S. Beavers, U.S. National Park Service, pers. comm., 2007). Three lines of evidence suggest that this foraging population has reached carrying capacity: (1) The significant increase in green turtle abundance throughout the archipelago over the last 10 to 20 years has been associated with a significant decrease in somatic growth rates at many foraging grounds around the Main Hawaiian Islands, including the west coast of the Big Island (Balazs and Chaloupka 2004), possibly the result of density dependence (Bjorndal et al. 2000); (2) Field measurements of body volume and mass as an index of body condition have shown that turtles at foraging locations near Kaloko have lower body condition indices than green turtles at other sites on the island of Hawai‘i (Kubis et al. 2008); and (3) recent necropsy reports cite emaciation as a probable contributor in the death of a number of green turtles found stranded at foraging locations along the west coast of the Big Island (or Kona coast) (Work 2007; Work 2008a; Work 2008b).  From the Tropics to the Poles, Wabnitz and Hoover  The trophic model is also intended as a ‘baseline’ of ecosystem state for Kaloko prior to major developments projected for areas around the park. Concern has been expressed over the future health of Kaloko’s coastal resources given proposed plans for the development of lands adjacent to the south boundary of the park, including a 300% expansion of a small-boat harbor, and construction of hotels, condominiums, and a light industrial park (Gibbs et al. 2007). Expected impacts include a reduction in groundwater flow, an important feature at Kaloko, with a concomitant increase in groundwater loads of sediment, nutrient, and chemical pollutants (Oki et al. 1999; Paytan et al. 2006; Knee et al. 2008; Johnson et al. 2008). The goals of this study were therefore threefold: (1) to develop an ecosystem model of the marine portion of the park to synthesize available data and describe the ecological structure and processes of the reef system at Kaloko; (2) to ascertain whether Kaloko green turtles are at carrying capacity, by determining whether grazing by green turtles and other reef herbivores matches overall algal production; and (3) to provide the management community with a tool that can simulate the effects of increased urban development in the Kaloko area, and compare the outcomes of a range of potential management scenarios.  MATERIALS AND METHODS Study Area Kaloko-Honokōhau National Historical Park was established in 1978 "to provide a center for the preservation, interpretation, and perpetuation of traditional native Hawaiian activities and culture, and to demonstrate historic land use patterns, as well as to provide a needed resource for the education, enjoyment, and appreciation of such traditional native Hawaiian activities and culture by local residents and visitors”. The legislative boundary of the park covers a total of 5.17 km2, 2.48 km2 of which are marine (Gibbs et al. 2007). The terrestrial portion encompasses a number of anchialine pools and wetland complexes, two large ponds modified for fish production by early Hawaiians, and a fish trap (Kaloko, Aimakapā and ‘Ai’ōpio respectively; figure 1). The coastal waters and reefs of Kaloko are within the West Hawaii Fisheries Management Area and managed by the State of Hawaii. The National Park Service and the State are currently discussing a joint management agreement (S. Beavers, pers. comm., 2009). The legislated park boundary extends offshore for about 1,000 m at the widest point and to maximum depths of ca. 70 m (Parrish et al. 1990). Past this boundary the seabed quickly drops off to depths > 180 m. Approximately 73% of the study area is hardbottom. The remaining 27% comprises unconsolidated sediment and artificial/historical features. Most of the hardbottom area has 10% to < 50% coral cover, and ~ 12% exhibits moderately high (50% to <90%) to high (90% to 100%) coral cover (Gibbs et al. 2007). Overall, benthic habitats are considered relatively healthy with no signs of diseased corals or non-native algal species (Marrack et al. 2009; Weijerman et al. 2009). Low sedimentation rates and the presence of relatively high coral cover in protected locations, suggest that currently, the reef habitat in Kaloko is primarily controlled by natural wave-induced stresses (DeVerse 2006). Figure 1: Map of Kaloko-Honokōhau National Historical Park (Kaloko).  72                  Baseline Trophic Relationships in Kaloko-Honokōkau, Hawai’i   Modeling Approach Ecopath and Ecosim (EwE) version 5.1.208 was used for all modelling work (Christensen et al. 2005, http://www.ecopath.org). The Ecopath component of any EwE model provides a quantitative representation of the studied ecosystem for a defined time period. In other words it is a snapshot of the resources in an ecosystem and their interactions, represented by trophically linked mass-balanced biomass ‘pools’ (Polovina 1984; Christensen and Pauly 1992). The biomass pools, hereafter referred to as functional groups, consist of a single species, or species groups representing ecological guilds. The basic idea behind the mass-balance approach is that “at any time within the system, and within the elements of that system, the amounts of matter that flow in must balance the amount that goes out plus the change in biomass” (Pauly and Christensen 2002 / p. 215). Ecopath therefore operates under two main assumptions: 1. That biological production within a functional group equals the sum of mortalities, i.e., on an annual basis, biomass and energy in an ecosystem are conserved (Walters et al. 1997; Walters and Martell 2004). This relationship can be expressed as follows:  where Bi and Bj are biomasses of prey (i) and predator (j) respectively; (P/B)i is the production to biomass ratio - equivalent to total mortality (Z) under most circumstances (Allen 1971); (Q/B)j is the food consumption per unit biomass of (j); DCji is the fraction of prey (i) in the average diet of predator (j); Yi is the total fishery catch rate of group (i); Ei is the net migration rate (emigration – immigration); BAi is the biomass accumulation rate of group (i); and EEi is the ecotrophic efficiency, defined as the fraction of production that is consumed within the system or caught by fishers; and 2. That consumption within a group equals the sum of production, respiration, and unassimilated foods. This relationship can be expressed as follows:   where GS is the proportion of unassimilated food; and TM is the trophic mode expressing the degree of heterotrophy of groups represented within the system, with 0 representing autotrophs, 1 heterotrophs, and intermediate values facultative consumers.  Ecopath then uses a set of algorithms to simultaneously solve n linear equations of the form of the first equation, where n is the number of functional groups. For each functional group, three of the basic parameters: Bi, (P/B)i, (Q/B)i  or EEi must be known, in addition to the fisheries yield (Yi) and the diet composition. Units of the model are expressed in t·km-2·year-1 wet weight organic matter for flows and t·km-2 for biomasses. Production per unit biomass (P/B) and consumption per unit biomass (Q/B) have the dimension year-1. For a review of EwE’s capabilities and limitations see Christensen and Walters (2004), Plaganyi and Butterworth (2004), and Plaganyi (2007). To balance the model, changes were first made to the diet matrix, as diet compositions represent only snapshots of the feeding habits of individual species and are likely to be relatively variable based on location and time periods of data collection. The model required only minor adjustments and was considered balanced when: (i) the model produced realistic ecotrophic efficiencies (EE<1); (ii) values of the production to consumption ratio (P/Q) for functional groups were between 0.05 and 0.35, with the exception of groups with fast growth rates (higher ratios), and top predators (lower values) (Christensen et al. 2005). )1()()/()/(1iiiiiiijjjjii EEBPBBAEYDCBQBBPB  GSBQBPTMQGSBPBBQB  )/()1()1()/()/(From the Tropics to the Poles, Wabnitz and Hoover  Model Parameters and Functional Groups The model represented an annual average situation of ecosystem conditions in the marine portion only (i.e., to the exclusion of the anchialine pools, fish ponds, and fish traps) of Kaloko in 2005. A total of 26 groups were defined, eight of which were fish, spanning the main trophic components of the ecosystem (including detritus; table 1, figure 2). The eight fish groups represent 106 species, recorded during underwater visual census (UVC) studies, which were primarily aggregated into functional groups based on ecological and biological similarities (e.g., diet, size, habitat, mortality) (Appendix 1).  Biomass estimates for individual species were based on values derived from field studies or from the literature. For species with data reported only for select habitats (e.g., fish), biomass values from these specific habitats were extrapolated to the entire park by calculating an area-weighted biomass for each species relative to the proportion each benthic habitat category covers within park waters, following Gibbs et al. (2007). For instances where P/B was equal to only natural mortality, estimates were taken directly from the literature or derived using the empirical formula of Pauly (1980). For exploited species, the fishing mortality was assumed to be proportional to M depending on the fishing pressure exerted (see below). Where possible, the consumption of each group was obtained through field studies; otherwise it was estimated from empirical equations such as those available in Fishbase (www.fishbase.org) for all finfish. The diet matrix was constructed using data from studies in Hawai’i (preferentially the Kona Coast); where no such data were available, the matrix was complemented with diet data obtained from the literature for the same species in similar ecosystems.  Fisheries Archaeological evidence reveals that seafood, particularly species encountered on coral reefs, was part of the traditional diet of the earliest human inhabitants of the Hawaiian Archipelago (Smith 1993). Kaloko’s waters were designated as a Fisheries Management Area (FMA) on December 31, 1999 (HAR §13-60.3-14), effectively banning collection of fish for the aquarium trade in park waters, in response to declines of TL B P/B Q/B EE P/Q P/R Catches Group name (t/km²) (/year) (/year) (t/km²/year)1 Spinner dolphins 3.21 2.7400 0.151 11.519 0.007 0.013 0.0172 Monk seals 3.89 0.1790 0.121 11.508 0.033 0.011 0.0133 Sea birds 3.17 0.0024 0.127 76.515 0.012 0.002 0.0024 Rays 3.15 4.2330 0.200 3.100 0.002 0.065 0.0655 Sharks and jacks 3.53 0.0700 1.058 5.100 0.453 0.207 0.350 0.030          6 Hawksbill sea turtles 3.18 0.0540 0.100 3.500 0.066 0.029 0.0297 Green sea turtles 2.00 1.5910 0.109 6.764 0.039 0.016 0.0218 Reef fishes - piscivores 3.39 1.7295 0.615 6.121 0.527 0.100 0.144 0.003          9 Reef fishes - herbivores 2.02 20.3350 1.400 27.149 0.205 0.052 0.069 0.162          10 Reef fishes - corallivores 2.60 0.5417 2.100 12.918 0.547 0.163 0.25511 Reef fishes - detritivores 2.00 2.2598 1.900 32.272 0.282 0.059 0.079 0.018          12 Reef fishes - MIF 3.13 9.7610 0.950 8.108 0.394 0.117 0.172 0.130          13 Reef fishes - SIF 2.84 0.5440 1.700 9.581 0.224 0.177 0.28514 Reef fishes - Zoo 2.85 3.0460 1.450 13.378 0.585 0.108 0.157 0.004          15 Urchins 2.00 280.0000 0.484 8.547 0.056 0.057 0.07616 Crown of thorns 2.59 0.1170 0.411 9.000 0.007 0.046 0.06117 Benthic Invertebrates 2.18 42.5381 2.910 15.250 0.950 0.191 0.31318 Corals 1.58 130.0000 0.140 2.100 0.594 0.067 0.07519 Octocoral 2.07 2.9000 0.200 4.630 0.484 0.043 0.05420 Macroalgae 1.00 22.6910 9.824 - 0.925 - -21 CCA 1.00 37.8180 1.770 - 0.358 - -22 Turf algae 1.00 128.7800 19.000 - 0.942 - -23 Turf algae_lava bench 1.00 3.0650 25.000 - 0.921 - -24 Zooplankton 2.02 1.2400 219.000 949.000 0.979 0.231 0.62525 Phytoplankton 1.00 3.2900 325.458 - 0.984 - -26 Detritus 1.00 100.0000 - - 0.694 - -Table 1 : Trophic parameters for all functional groups of the balanced Kaloko model. Outputs from the balanced model are presented in bold. B = biomass; TL= Trophic level; P/B= Productivity biomass ratio; Q/B = consumption to biomass ratio; EE= Ecotrophic efficiency; P/Q Production to consumption ratio or gross efficiency; P/R= Production to respiration ratio. MIF = Mobile Invertebrate Feeders; SIF = Sessile Invertebrate Feeders; Zoo = Zooplanktivorous fish; CCA = Crustose Coralline Algae 74                  Baseline Trophic Relationships in Kaloko-Honokōkau, Hawai’i   species targeted by collectors (Tissot et al. 2004). Baseline surveys conducted prior to FMA closure, comparing aquarium collection sites (including Kaloko) and ‘control sites’, showed a significant depletion in aquarium species compared to non-target species. Surveys conducted subsequent to the FMA designation recorded significant increases in the overall abundance of fish targeted by collectors (Tissot et al. 2004). However, at Kaloko specifically, post-closure surveys showed no significant change in the abundance of yellow tangs (Zebrasoma flavescens), the most sought-after species. The lack of response may have been due to the small size of the Kaloko FMA and the relatively small amount of suitable habitat compared to other FMAs included in the analysis (Hoover and Gold 2005). In the years since the surveys, biomass may well have responded to the closure, and because fishing pressure has been removed, P/B was set to M for all formerly collected species.    From a food-fish perspective, the park is a popular location for subsistence fishing and shoreline gathering, traditional activities that are permitted as long as they are consistent with state law and park mandates (i.e., with legal fishing gear for personal consumption DeVerse 2006). Harvesting is done primarily from shore using a variety of methods, such as throw nets, spear, and pole fishing. Gill, or “lay” netting, a serious threat to marine resources including marine mammals and sea turtles, was restricted within park waters in August 2005 to locally constructed, handmade nets of natural fibers. The state of Hawai‘i does not have recreational and subsistence permitting or reporting requirements (Friedlander and Parrish 1997), despite surveys in the late 1980s indicated that 19-35% of human residents fish (Smith 1993) and recent studies concluding that total take of coral reef fish species by recreational fishers is far greater than commercial catches (Zeller et al. 2005), with shore-based fishing representing a much larger part of total fishery take than boat-based fishing for nearly all targeted species (Williams et al. 2008). Results from a recent study show that fish biomass of several target species groups was negatively correlated with local human population density and concluded that fishing was the prime driver of those trends (Williams et al. 2008). Of the 18 coral reef locations included in the study for comparison and widely spread throughout the Main Hawaiian Islands, none were located on the west coast of Hawai‘i Island. Tissot et al. (2004) and Tissot et al. (2009) reported on the effectiveness of community-based co-management and an MPA network along the western Kohala-Kona Coast of Hawai‘i Island as means to increase productivity of species targeted for the aquarium trade specifically. However, data are lacking to 4321Spinner dolphinsMonk sealsSea birdsRaysSharks and jacksHawksbillsGreen sea turtlesPiscivoresHerbivoresCorallivoresDetritivoresMIFSIF ZooUrchinsCrown of thornsBenthic InvtsCoralsOctocoralMacroalgae CCA Turf algae Turf algae_LBZooplanktonPhytoplankton DetritusFigure 2 : Graphical representation of trophic flows within the Kaloko reef ecosystem. Each functional group is identified here by an illustration (© M. Bailey); where relevant, an image of a species representative of its guild is depicted. Images are not drawn to scale or proportional to the group’s biomass. The light grey horizontal lines and associated numbers represent trophic levels; lines connecting individual functional groups represent trophic links. Zoo = Zooplanktivorous fish; MIF = Mobile Invertebrate Feeding fish; SIF = Sessile Invertebrate Feeding fish; Turf algae_LB = turf growing on the lava bench area; CCA = Crustose Coralline Algae; Benthic Invts = Benthic invertebrates From the Tropics to the Poles, Wabnitz and Hoover  quantitatively assess patterns and trends relating to species targeted by commercial as well as recreational and subsistence fisheries along the western shores of the Big Island.  Hawai‘i’s Department of Aquatic Resources (DAR) conducts regular surveys at 14 sites in West Hawai‘i, including one within Kaloko adjacent to the Honokōhau harbor, focusing on ‘resource fish’ (DAR unpublished data, in Weijerman et al. 2009). Results show that among these 14 sites, the one located in Kaloko has the lowest biomass. However, as this site is located just outside of the harbor, this low biomass may be due partly to its accessibility to spear fishers. More generally fish in park waters do not appear as grossly depleted as in some coastal waters of the state (Parrish et al. 1990; Beets et al. 2010, S. Beavers, pers. comm., 2008).  Of the few existing case studies documenting catch levels, results show landings to be either representative of fishers targeting specific species, or proportional to species’ biomass on the reef. In 1997, Friedlander and Parrish (1997) measured standing stock and quantitatively estimated catch and effort of a small recreational/subsistence fishery at Hanalei Bay, Kauai, Hawai‘i. In the absence of catch or effort data specific to Kaloko, it was assumed that catch data to standing stock proportions in the park would be comparable to those in Hanalei Bay (Friedlander and Parrish 1997); with values adjusted to reflect species known to be targeted at Kaloko. For example, although corallivores featured in the catch at Hanalei Bay, they are not target species at Kaloko. As no data were available on macroinvertebrates in Kaloko (with the exception of urchins), nor information regarding whether fishing occurs on those groups in the park, no catch was allocated to the ‘benthic invertebrates’ functional group.  The existing fisheries were divided into two ‘fleets’, one targeting sharks and jacks specifically and a recreational/subsistence fishery representing fishers operating mainly from shore using pole, spear, and line, and targeting small reef fish. The ratio of catch to standing stock was calculated for fish families targeted by fishers from Friedlander and Parrish (1997), and doubled given that fishing pressure is likely to be higher now than it was in 1997 (DAR unpublished data in Weijerman et al. 2009). In the absence of other information these indices were applied to fish families known to be harvested in Kaloko (E. Brown, unpublished data). For species groups targeted in Kaloko, but for which data were not available from the Friedlander and Parrish (1997) study, the same values were applied as for fish families in the same functional group. Fishing mortality for functional groups, and by extension individual species, was subsequently calculated using F = catch/biomass.  Functional Groups Spinner dolphins Hawaiian spinner dolphins belong to a stock that is separate from those involved in the tuna purse-seine fishery in the eastern tropical Pacific (NOAA 2005a). A 2002 shipboard line-transect survey of the entire Hawaiian Islands Exclusive Economic Zone resulted in an abundance estimate of 2,805 spinner dolphins (Barlow 2006). Only a fraction of this number is likely to be considered resident within Kaloko waters and to feed there regularly. Norris (1994), in his intensive study of spinner dolphins of the Kona Coast of Hawai‛i, suggests that the waters surrounding this island may have a large, relatively stable ‘resident’ population with a minimum of 960 animals regularly frequenting the shore of the Big Island. Based on sightings information provided by National Park Service staff, an average of one hundred 68-kg dolphins were assumed to utilize park waters on a regular basis, a biomass of 2.74 t·km-2.  The P/B value was assumed to be equal to M (natural mortality) and was derived from a life history table model (Barlow and Boveng 1991) that estimates survivorship and mortality according to the longevity of a given species. Spinner dolphins live on average to be 20 years old (NOAA 2003; NOAA 2005a) and P/B was therefore set at 0.151 year-1. Consumption per unit of biomass (Q/B) was first calculated using an empirical equation for daily ration R=0.1·W0.8, as modified from Innes et al. (1987) in Trites and Heise (1996), where W is body weight in kg and R the daily ration in kg·day-1 (Method 1). Hunt et al. (2000) describe energy requirements using the equation E = a·W0.75 where E is the energy requirement per day (kcal·day-1), W the mean body weight (kg) and a is a coefficient varying with the group of mammals (a=320 for otariids, 200 for phocids, 192 for mysticetes, 317 for odontocetes, and 320 for sea otters). The coefficient of 0.75 is intended to apply to mammals in general and was changed to 0.714 following Hunter (2005) who estimated a more precise 76                  Baseline Trophic Relationships in Kaloko-Honokōkau, Hawai’i   coefficient for marine mammals (Method 2). Benoit-Bird (2004) found the daily maintenance energy needs of a spinner dolphin to range between 2,430 kcal and 4,050 kcal (measured on a stranded animal and large adult males respectively), with an estimate of 3,520 kcal for an average adult (Method 3). Using an average weight of 68 kg for an individual spinner dolphin resulted in Q/B estimates of 15.70 year-1, 12.2 year-1, and 6.66 year-1 for methods 1, 2, and 3 respectively. The average of 11.519 year-1 was used.  Spinner dolphins in Hawai‘i have been observed to rest during the day and feed at night, as many of their prey species, small mesopelagics (< 20 cm), are organisms that rise from deep water to the surface near dusk (Norris and Dohl 1980; Norris et al. 1994; Benoit-Bird and Au 2003). The Hawaiian spinner populations are also known to take bottom dwelling, and small numbers of surface dwelling species, as well as crustaceans (Perrin and Gilpatrick 1994; Würsig et al. 1994; Perrin 1998). Monk Seals Counts at the six main Northwest Hawaiian Islands (NWHI) subpopulations of monk seal (Monachus schauinslandi), an extrapolation of counts at Necker and Nihoa Islands, and counts at the Main Hawaiian Islands (MHI) led to a best estimate of a total population size for Hawai‘i of 1,202 individuals in 2006 (NMFS 2007). A 2001 aerial survey determined a minimum abundance of 52 individuals in the MHI and remains the most recent available estimate (Baker and Johanos 2004). However, occasional sighting reports from the area may not reflect the actual rate of park shoreline habitat use by monk seals. Rather, they represent the minimum shoreline usage because sightings are opportunistically collected and heavily biased by reporting effort. From existing data, monk seals found to haul out in the park cannot be assumed to utilize the parks' near shore resources as foraging habitat (i.e., have the resources contribute to their diet). Thus, although these animals may utilize the park's waters for foraging, they may well derive a greater portion of their diet from surrounding waters. In light of the paucity of data concerning monk seal usage of the area, our biomass estimate included two individuals (partly based on data presented by Baker and Johanos (2004)) of an average mass of 187.5 kg (NOAA 2005b; NMFS 2006) (i.e., B=0.179 t·km-2) (T. Wurth, Pacific Islands Fisheries Science Center, pers. comm., 2007). These values are meant to serve as a placeholder until more information becomes available. The net productivity rate for monk seals, based on overall declines of this species in Hawai‘i, is currently assumed to be -0.019 year-1, although stock assessment data highlight that population trends vary considerably among the six main subpopulations (Baker and Johanos 2004). Although the MHI monk seal population may be on the rise (Baker and Johanos 2004), this increase remains unconfirmed and abundance estimates appear to be too low to strongly influence current total stock trends. Application of the MHI overall population productivity trend, Barlow and Boveng’s (1991) life history table model, and a longevity estimate of 25 years, led to a P/B estimate of 0.121 year-1. Consumption per unit of biomass was calculated according to methods 1 and 2 introduced above for spinner dolphins. Using an average monk seal mass of 187.5 kg, and two estimates of average caloric prey-content calculated from data presented in Goodman-Lowe et al. (1999a) and Goodman-Lowe et al. (1999b), resulted in Q/B values ranging from 10.27 year-1 to 12.82 year-1, with the average of 11.508 year-1 being used for the model. These values are much lower than the estimate of Polovina (1984) or Massicot (2006), who assumed, respectively, that monk seals must consume, on average, 45 or 36.5 times their weight in food per year.  Based on the analysis of identifiable hard parts found in regurgitate and fecal material, Goodman-Lowe (1998) reported that fish comprised the grestest proportion of monk seal diets, with typical prey species including marine eels (Congridae, Muraenidae, and Ophicthidae), and various reef fish such as wrasse (Labridae), squirrelfish and soldierfish (Holocentridae), as well as triggerfish (Balistidae). The remainder of the diet was comprised of cephalopods and crustaceans (Longenecker et al. 2006).  Birds Most Hawaiian lowland waterbirds, thought here to include waterfowl, rails, shorebirds, and waders (K. Uyehara, pers. comm., 2007), tend to be found in and around traditional fishponds (Morin 1994), which are not included in this model. Two of the most important fishponds along the Kona Coast, 'Aimakapa and Kaloko, lie within park boundaries. In addition to fishponds, the park offers a range of habitats, including From the Tropics to the Poles, Wabnitz and Hoover  sandy and rocky intertidal shoreline, anchialine pools, brackish water wetlands with some mudflats, coastal strand vegetation, all of which are known to attract a wide variety of bird species (Morin 1994).  The only indigenous resident birds remaining in Kaloko-Honokōhau are three endemic, endangered waterbird species: the Hawaiian stilt (Himuntopus mexicanus knudseni), the Hawaiian coot (Fulica alai), and the indigenous black-crowned night heron (Nycticorax nycticorax hoactli). Despite predominantly utilizing fishpond habitats, the endangered Hawaiian stilts, the black-crowned night-herons, and all of the waterbirds use the rocky intertidal beach areas for feeding, especially during low tides (Morin 1994). This usage was also confirmed by K. Uyehara, S. Waddington, and S. Beavers (pers. comm., 2008), and therefore the Hawaiian stilt, sanderling (Calidris alba), ruddy turnstone (Arenaria interpres), wandering tattler (Heteroscelus incanus), Pacific golden-plover (Pluvialis fulva), and black-crowned night-heron were included in the Kaloko model.  Biomass estimates were based on sighting surveys conducted in an area of about 0.02 km-2, and these were subsequently extrapolated to the 0.1 km-2 portion of the park available for birds to forage in (S. Waddington, Cyanotech Corporation, pers. comm., 2007). Individual species’ weight data were extracted from a number of sources (e.g., Anonymous 1996a; Anonymous 1996b; Reed et al. 1998; Nettleship 2000; Gill et al. 2002; MacWhirter et al. 2002). Biomass calculations were weighted by individual species biomass’ contribution within the group and total bird biomass was estimated to be 0.0024 t·km-2. Very little information was found to guide estimates of P/B values for waterbirds found in Kaloko. Polovina (1984) used a value of 5.4 year-1 for birds in French Frigate Shoals (FFS). However, bird species at FFS are mostly pelagic seabirds, which typically suffer higher mortality rates than the coastal waterbirds encountered in Kaloko. Hence the overall P/B value used here (0.127 year-1) was computed from survivorship rates found in a number of publications (e.g., Reed et al. 1998; Nettleship 2000; Gill et al. 2002) and weighted by individual species’ biomass contribution to the overall group. Q/B was determined by first calculating the ration for each species using the empirical formula derived by Nilsson and Nilsson (1976, in Wada 1996):  where R is the daily ration in g per day, and W is body weight also in g. The values were then averaged across species based on biomass contributions by individual species and resulted in a group Q/B estimate of 76.515 year-1.  Hawaiian stilts are opportunistic feeders that eat a variety of invertebrates and vertebrates found in shallow water and mudflats, such as polychaete worms, small crabs, aquatic insects, and small fish (Mitchell et al. 2005; U.S. Fish and Wildlife Service 2005). Sanderlings’ diet are known to markedly change between seasons, consisting almost exclusively of insects during the breeding season, and small crabs, isopods, insects, amphipods, polychaetes, and small mollusks in winter (Perez Hurtado et al. 1997; Tsipoura and Burger 1999; Petracci 2002; Anonymous 2005a; Nuka et al. 2005). Pacific golden plovers feed primarily on terrestrial insects, but are also known to forage in the intertidal areas and opportunistically prey on aquatic invertebrates (Kato et al. 2000; Anonymous 2005b). Outside of the breeding season, ruddy turnstones are known to prey on crustaceans, mollusks, polychaetes, and small fish (Tsipoura and Burger 1999; Nettleship 2000; Anonymous 2005c). The diet of wandering tattlers varies with season and in winter tends to consist of invertebrates such as marine worms, aquatic insects, mollusks, crustaceans, and small fish (Gill et al. 2002; Anonymous 2005d). The black-crowned night heron is an opportunistic feeder, whose diet consists mainly of fish, though it will occasionally feed on other items such as earthworms, and aquatic and terrestrial insects (Wolford and Boag 1971). It has also been observed to feed on crayfish, mussels, squid, amphibians, lizards, snakes, and plant material (Davis 1993). Rays  The spotted eagle ray (Aetobatus narinari) and the manta ray (Manta birostris) are both known to transit through park waters. Manta rays are the focus of commercial night diving and research activities in and near Kaloko waters. Relatively few quantitative data are available for this functional group within the park. A biomass estimate of 4.233 t·km-2 was assigned to the group overall, based on the subjective species level WR log85.0293.0log 78                  Baseline Trophic Relationships in Kaloko-Honokōkau, Hawai’i   abundance rankings and use of park waters for feeding activities, as well as estimates of individuals’ biomass provided by T. Clark (University of Hawai‘i, unpublished data). The group’s P/B ratio (0.2 year-1) was estimated by averaging individual species’ natural mortality rate using Pauly’s formula (1980). Very little is known about foraging patterns and rates of mantas or spotted eagle rays. Our estimate of Q/B (3.1 year-1) was based on values for individual species derived from Fishbase as well as parameters presented in Olson and Watters (2003).  Rays as a functional group were assumed to feed on benthic invertebrates as well as zooplankton (Olson and Watters 2003; Wetherbee and Cortes 2004). Manta rays are known to mostly feed outside of park waters, but spotted eagle rays have been seen to regularly forage in the Honokōhau channel (T. Clark, unpublished data). Sharks and Jacks  Sharks are sighted offshore relatively frequently (S. Beavers, unpublished data), and in recent years tiger sharks have been repeatedly sighted within, and near the mouth of Honokōhau harbor (DLNR 2001; Thompson 2005; Meyers et al. 2009). Based on park data, tiger sharks (Galeocerdo cuvier) and whitetip reef sharks (Triaenodon obesus), as well as other top predators such as bluefin trevally (Caranx melampygus), bigeye trevally (Caranx sexfasciatus), mackerel scad (Decapterus macarellus), golden trevally, (Gnathanodon speciosus), doublespotted queenfish (Scomberoides lysan), bigeye scad (Selar crumenophthalmus) and greater amberjack (Seriola dumerili) are only rarely spotted in Kaloko waters. A biomass estimate of 0.07 tkm-2 was assigned to the group overall, based on findings in Friedlander and DeMartini (2002), the subjective species level abundance rankings offered on the park’s website (http://www.nps.gov/kaho/), and by Parrish et al. (1990).  The group’s P/B ratio (1.058 year-1) was estimated by averaging individual species natural mortality rates using Pauly’s formula (1980) and fishing mortality estimates based on Friedlander and Parrish (1997). Consumption rate was estimated at 5.1 year-1 and represents the average of values derived for individual species based on data derived from Fishbase.  In Hawai‘i, tiger sharks have a broad diet. As sharks increase in size, prey diversity and frequency of occurrence of large prey items increase. Teleost fish make up a large proportion of the diet of all size classes of sharks, and marine mammals and sea turtles are relatively uncommon, even in large sharks (Lowe et al. 1996). For most other species, diet data was derived from Fishbase. Hawksbill turtles Individually identified hawksbills are seen at specific sites within Kaloko, on a regular basis (S. Beavers, unpublished data). Although a number of them just travel through, hawksbill turtles have been filmed feeding and attempting to mate in park waters. It was assumed for three turtles to be ‘resident’ in Kaloko (G. Balazs and S. Beavers, unpublished data). Each turtle was assumed to weigh about 45 kg, resulting in a total biomass of 0.054 t·km-2.  Few estimates of hawksbill survival rates exist, thus the P/B value was based on Crouse (1999) and set to 0.109 year-1. As no data were available regarding foraging rates at Kaloko, the Q/B was set to be equal to 3.5 year-1.  Hawksbill turtles primarily feed on sponges and benthic invertebrates (Bjorndal 1997). In Hawai‘i, sponges are not abundant, and limited information gained through necropsies and visual observations indicate that hawksbill turtles appear to feed on sea cucumbers, fireworms, and red algae (S. Beavers and S. Hargrove, unpublished data). Green Turtles Data on green turtle sightings and measurements were obtained from the NOAA National Marine Fisheries Service's (NMFS) Marine Turtle Research Program (G. Balazs and S. Hargrove, unpublished data), and the National Park Service (NPS) (S. Beavers, unpublished data). The sum of the maximum mass for all 58 individual juvenile turtles caught in 2005 amounted to a total biomass of 1,600 kg. Turtles that were captured more than once in a year were only counted once and maximum mass for that individual was used in calculations. In instances where specific mass data were not available the following From the Tropics to the Poles, Wabnitz and Hoover  assumptions were made: (i) where size was available, mass was calculated using a length-weight relationship based on previous captures; and (ii) where size was not available, individuals were given an average value as derived from all turtles measured that year. Based on recapture rates since 2003, biomass outside the sampling area was estimated at ~25% of the biomass of individuals tagged within the study area, leading to a total green turtle biomass estimate of 0.806 t·km-2 within Kaloko. However, NPS resighting data suggest that the number of turtles tagged under the NMFS long-term turtle monitoring program is an underestimate of the total number of individuals that regularly use park waters. Based on weekly surveys and NPS mark and recapture data, 143-161 out of 196 turtles associated with Kaloko show high site fidelity, and can thus be considered Kaloko ‘resident’ (S. Beavers, NPS unpublished data). Using the average green turtle weight in 2005 of 27.6 kg, which is not significantly different from the average for all other years, led to an average green turtle biomass for Kaloko of 1.591 t·km-2. The model was run with this latter estimate.  In Hawai‘i, green turtles have benefited from effective protection under the U.S. Endangered Species Act since 1978 (43 Federal register 32808). It therefore seems fair to assume that natural mortality rate represents a good estimate of green turtles’ P/B ratio. Bjorndal et al. (2003) calculated true annual survival probability for green turtles protected from human induced mortality at Union Creek, Bahamas, to be equal to 0.891 year-1. Although predation rates are likely to be higher in Hawai‘i due to a greater natural abundance of tiger sharks (Witzell 1987), green turtles at Kaloko have generally been observed to show high site fidelity and not to migrate much. Indeed, individuals spend a significant portion of daytime hours, when tide height is sufficiently high, in very shallow waters or in the intertidal zone foraging on a dense mat of turf algae growing on a ‘lava bench’ (G. Balazs, pers. comm., 2008 and S. Beavers, NPS unpublished telemetry data). Therefore, the higher estimate, derived by Bjorndal et al. (2003), was used here (P/B=0.109 year-1).  No estimates of consumption rates were available for the resident green turtle population at Kaloko and, overall, very few data are available on natural consumption rates of green turtles, particularly for algal environments. Weekly surveys conducted by NPS staff do however indicate that the primary activity of turtles observed is feeding (S. Beavers, NPS unpublished data). The Q/B ratio for green turtles was estimated at 6.764 year-1 based on an average body mass intake of 1.8% per day. This proportion was recalculated from Brand et al. (1999) using the length weight relationship in Arthur et al. (2006), as the original study was conducted in Australia. It should be noted that both of these studies were conducted in seagrass environments. Our estimate is within the lower range of consumption rates estimated from seagrass intake rates in the Caribbean (Bjorndal et al. 2000). Data from a recent study conducted off the coast of Colombia suggest that consumption rates can be much lower (i.e., 1.9 year-1 (Amorocho and Reina 2008)). However, Amorocho and Reina’s (2008) findings were based on an experimental set up that involved moving animals into the laboratory every three days to be fed the equivalent of 1-2% body mass, a ratio considered to be a maintenance diet (Higgins 2003). Therefore, these consumption estimates may not be consistent with feeding rates observed in the wild. Over 275 species of marine algae and two seagrass species have been reported from green turtle crop and stomach samples in the Hawaiian Islands (Balazs 1980; Balazs et al. 1987; Russell et al. 2003; Russell and Balazs 2009). A study examining diets of immature green turtles at seven sites in the MHI showed that despite variations among foraging grounds, overall, individuals’ diets were dominated by red algae, with Acanthophora spicifera (an introduced species), Hypnea sp., Pterocladiella sp., and Cladophora sp. being prominent (Arthur and Balazs 2008). In the same study, all turtles appeared to have a base diet of algal turf, enhanced with desirable monogeneric stands when available. Hawaiian ‘turf algae’ are comprised of multiple species of compact, often filamentous, red and green algae with a canopy height of only a few millimeters. Given the high cover of turf algae and the low availability of macroalgae at Kaloko (Marrack et al. 2009; Weijerman et al. 2009), turf algae constitutes the primary dietary component of green turtles within park waters (G. Balazs and S. Beavers, pers. comm., 2009).  Reef Fish The Park Certified Species list comprises 203 species of reef fish, not all common however (NPS unpublished data), making Kaloko host to the greatest number of fish species of the four national parks surveyed in Hawai‘i (Hobson 1974). Beets et al. (2010) collected UVC transect data for three habitat classes: unconsolidated sediment [sand] (UCS), colonized hard bottom (CHB), uncolonized hardbottom (UCH) from mean low water to 30-m depth. Unconsolidated sediments were excluded from the analysis. 80                  Baseline Trophic Relationships in Kaloko-Honokōkau, Hawai’i   Although it represents one of the largest habitat classes in the study area, much of it is probably below 30 m depth, and it is unclear what proportion occurs at shallower depth, precluding reliable extrapolations from transect data to the entire park. In any event, few fish were recorded along transects on UCS and its exclusion therefore had a negligible effect on overall biomass estimates. A single species, Malacanthus brevirostris, was documented to occur only on UCS. Beets et al.’s (2010) fish biomass values for CHB and UCH were extrapolated to the entire park by calculating an area-weighted biomass for each species relative to the proportion of each benthic habitat category within park waters. According to Gibbs et al. (2007) CHB represents 60.73% and UCH 11.74% of park substrate. To facilitate data comparisons with ongoing monitoring studies, and to ensure greater transferability and relevance of results to existing efforts on the ground, fish species were grouped according to the same functional groups used by Beets et al. (2010). These were corallivores, detritivores, herbivores, mobile invertebrate feeders (MIF), piscivores, sessile invertebrate feeders (SIF), and zooplanktivores (Zoo) (see Appendix 1 for species list).  P/B rates were derived from natural mortality estimates, based on Pauly (1980), and fishing mortality estimates based on the catch data outlined in the fisheries section above. Functional group estimates were weighted by individual species biomass contribution within their respective group. Q/B rates for each functional group were estimated based on information derived from Fishbase for individual species and using diet information (outlined below) to classify individual species as herbivores or detritivores. As was done for P/B values, Q/B estimates for a given guild were weighted according to individual species’ biomass contributions within their respective groups.  A diet matrix was developed from a detailed analysis of food habits of some teleost fish along the Kona Coast (Hobson 1974) as well as from information gleaned from Fishbase and a large number of published studies (e.g. Bruggemann et al. 1994; Guiasu and Winterbottom 1998; Choat et al. 2002; DeFelice and Parrish 2003; Paddack et al. 2006; Dierking et al. 2009). Recent evidence from studies conducted in Australia shows that fish nominally considered as herbivores may actually feed on a number of items other than algae, especially detritus, and that, as such, explicitly herbivorous taxa are a minority (Crossman et al. 2001; Choat et al. 2002; Choat et al. 2004; Crossman et al. 2005). For continuity and comparative purposes the same functional groups as those used for monitoring activities in Hawai‘i were retained, but for the relevant species these new findings were taken into consideration when assembling the diet composition matrix.   Urchins  Sea urchins are highly abundant at Kaloko and were examined separately from benthic invertebrates in more general terms due to their substantial contribution to grazing. Biomass estimates were based on surveys conducted at 10 m depth in 2006 by Marrack et al. (2009) and Weijerman et al. (2009). The most frequently encountered urchins during surveys were Echinometra mathaei, Echinothrix spp. (i.e., Echinothrix diadema and Echinothrix calamaris), Heterocentrotus mammilatus and Tripneustes gratilla. Test size for Echinothrix spp., T. gratilla, E. mathaei, and H. mammilatus were recorded on the reef by local researchers (M. Weijerman, U.S. National Park Service, pers. comm., 2008; H. Jessop, University of Hawai‘i, pers. comm., 2008) and converted to biomass based on published test size-weight relationships (Dotan 1990; McClanahan and Kurtis 1991; Rahman et al. 2001; Rahman et al. 2004; Muthiga and Jaccarini 2005). Biomass survey data were extrapolated to the whole park by calculating an area-weighted biomass for all species based on the assumption that urchins chiefly occur on colonized substrate with at least 10% coral cover (M. Weijerman, pers. comm., 2008). Total biomass was reduced from an estimated total of 294 t·year-1 to 280 t·year-1 to account for likely reduced densities of urchins in deeper areas of the park compared to surveyed depths (F. Parrish, NOAA, National Marine Fisheries Service, Pacific Islands Fisheries Science Center, pers. comm., 2008). P/B rates were calculated for each species using the functional relationships for mortality developed by Brey (2001). The group’s overall P/B rate was weighted according to individual species’ biomass contributions to the guild and estimated at 0.484 year-1. Q/B estimates for T. gratilla were based on feeding experiments conducted on T. gratilla using averages for three species of algae employed in trials (M. Deagle, University of Hawai‘i, pers. comm., 2007), and on T. gratilla and Echinothrix sp. using Gracilaria only (H. Jessop, pers. comm., 2007). Values of 14.72 year-1 and 13.9 year-1 for these two studies respectively, corroborate the findings of Stimson et al. (2007) who From the Tropics to the Poles, Wabnitz and Hoover  conducted food preference trials on T. gratilla. As no data were available for E. mathaei for Hawai‘i, Q/B for this species was set at 4.44 year-1 based on data from Carreiro-Silva and McClanahan (2001) and McClanahan and Kurtis (1991). This value is much lower than the estimate of 16.51 year-1 derived by Hiratsuka and Uehara (2007) from laboratory feeding trials. Their estimate is based on feeding sea urchins ad libitum a diet prepared from turf algae and agar over a 7-day period. For Echinothrix sp. the Q/B was set at 7.86 year-1. This represents the average of values reported by Carreiro-Silva and McClanahan (2001) and those obtained from feeding trials by H. Jessop (pers. comm., 2007). Consumption rates determined under laboratory conditions are frequently higher than those derived under natural conditions. In the absence of data for H. mammilatus, this species was assigned the same Q/B rate as Echinothrix sp. (intermediate between E. mathaei and T. gratilla). The guild’s overall Q/B of 8.547 year-1 was calculated by weighting individual species’ Q/B by their respective biomass contributions.  Although echinoids can show feeding preferences (de Ridder and Lawrence 1982), they are typically opportunistic feeders with their diets varying according to habitat and season. Observations at Kaneohe Bay, Oahu, Hawai‘i, by Stimson et al. (2007) show that T. gratilla’s diet composition typically reflects reef algal distribution, with individuals observed feeding on a variety of macroscopic algae, coralline algae, endolithic algae, and turfs. Based on these findings, the proportion of different algae groups encountered at Kaloko was applied to T. gratilla’s diet. In Hawai‘i, E. calamaris has been observed feeding on coralline algae, filamentous algae, brown algae (Castro, 1971 in de Ridder and Lawrence 1982). Echinothrix diadema is known to forage on algae and encrusting organisms (Mortensen, 1940 in de Ridder and Lawrence 1982). Heterocentrotus mammilatus has been seen to graze algae from bare substrate or the coral surface (Mortensen, 1943b and Dart, 1972 in de Ridder and Lawrence 1982). Although a number of studies have looked at the palatability and consumption rates of different algal species by sea urchins in laboratory conditions, few have looked at the proportion of macrophytes in sea urchins’ guts from the field in Hawai‘i. It was therefore assumed that Echinothrix sp. and H. mammilatus had diets comparable to that of T. gratilla, but the proportion of crustose coralline algae (CCA) was increased in the latter species based on the findings above and those by Regis and Thomassin (1983). Echinometra mathaei is a generalized herbivore, feeding on a variety of macrophytes (McClanahan and Muthiga 2001), and preferentially on turf growing on the surface of dead coral or pavement, which explains why calcium carbonate sediments are usually the largest fraction of the gut content of Echinometra (Odum and Odum 1955; McClanahan and Kurtis 1991). These findings were corroborated by results in Black et al. (1984) and Mills et al. (2000) who found that inorganic material constituted 73% of gut contents.  Crown of Thorns Starfish The crown of thorns starfish (Acanthaster planci) is a large, predatory asteroid, which feeds almost exclusively on scleractinian corals. It affects coral reef communities to a far greater extent than most other species of coral reef animals (Birkeland and Lucas 1990). Population outbreaks of A. planci throughout the Pacific have raised concern amongst both the public and scientific communities due to the resultant extensive coral mortality and subsequent long-lasting impacts on the health of the coral reef community (Pearson 1981; Moran 1986). As a consequence of the potential impacts of this corallivore, A. planci was included here as a separate functional group.  Based on surveys conducted at 10 m depth in 2006, both Marrack et al. (2009) and Weijerman et al. (2009) recorded a total of two individuals on all transects within Kaloko. Assuming that an individual weighs on average 466 g (Branham et al. 1971), and that A. planci only occurs in areas with > 50% coral cover, a biomass estimate of 0.117 t·km-2 was derived for Kaloko.  The P/B ratio was calculated based on Brey’s (2001) linear regression where maximum age was set to eight years (Zann et al. 1990) and maximum weight was derived from Branham et al. (1971). This led to an estimate of 0.411 year-1.  The Q/B ratio was derived based on information provided in Moran (1990), Keesing and Lucas (1992), Reyes-Bonilla and Caldero-Aguilera (1999), and Scandol (1999) and ranged between 5.969 year-1 and 12.065 year-1. An average of 9 year-1 was used here.  Adult A. planci feed mainly on hermatypic scleractinian corals, and preferentially on acroporids (Birkeland and Lucas 1990). 82                  Baseline Trophic Relationships in Kaloko-Honokōkau, Hawai’i   Benthic Invertebrates Uncertainty in estimates of invertebrate biomass is likely, as these groups are the least studied and the most under-represented in the coral reef literature. Very little qualitative or quantitative data are available on benthic invertebrates for the island of Hawai‘i. The cryptic nature and nocturnal patterns of many benthic invertebrates, as well as the rugosity of benthic cover on coral reefs, combine to make accurate sampling of many benthic invertebrates extremely difficult (Klumpp and Pulfrich 1989; Sorokin 1993). Based on the national park’s website, the majority of species are listed as uncommon to rare, with the exception of the white-spotted cucumber (Actinopyga mauritiana), snapping shrimp (Alpheus crassimanus), the helmet urchin (Colobocentrotus atratus), and the lightfoot crab (Grapsus tenuicrustatus), which are all listed as common. In surveys conducted two decades ago, Parrish et al. (1990) identified a number of invertebrates in the park, but listed only a few species as common: the lightfoot crab, sea cucumbers (Holothuria sp.), the black purse shell (Isognomon californicum), black nerite (Nerita picea) (freshwater, brackish), banded coral shrimp (Stenopus hispidus), nerita snail (Theodoxus neglectus) (freshwater, brackish), white spotted sea cucumber (Actinopyga mauritiana), helmet urchin, and the cushion star (Culcita novaeguineae). Based on Parrish et al. (1990), the largest and most conspicuous invertebrates at Kaloko were sea urchins (echinoids) (dealt with here as a separate group, see above), followed by sea cucumbers (especially Actinopyga mauritiana and Holothuria atra), and sea stars. The authors also reported that ophiuroids are probably abundant, but that due to their nocturnal behavior, they are probably underestimated by census data, while a few species of mollusks (e.g., Isognomom californicum, Theodoxus neglectus, Nerita picea) and crustaceans (most notably the lightfoot crab Grapsus tenuicrustatus) were widespread and common to abundant (as per Parrish et al. 1990) in the intertidal. Although not surveyed, it seems reasonable to assume that polychaetes would constitute a significant proportion of the benthic invertebrate functional group. Benthic invertebrates overall constitute an important source of food for some fish and large invertebrates (Hobson 1974). Given the absence of quantitative data for benthic invertebrates in Kaloko, Ecopath was allowed to estimate the biomass of this group for all model runs based on the assumption that EE ~0.95.  In the absence of more detailed information, the P/B and Q/B ratios were adopted from Tudman (2001) and estimated at 2.910 year-1 and 15.250 year-1 respectively. The diet of this group was set to consist chiefly of zooplankton, phytoplankton, and detritus (Brey 2001). Corals In two benthic habitat studies for Kaloko, average coral cover at 10 m depth ranged between 30.70% ±8.13% and 70.10% ±6.64% at the northern end (Marrack et al. 2009), and 31.4% ±7.4% to 58.3% ±5.5% in the southern portion of the park (Weijerman et al. 2009). Both these studies targeted areas with some coral cover (Marrack et al. 2009; Weijerman et al. 2009). Wide variability in coral cover depending on location within park waters was also reported by Gibbs et al. (2007) who found that 1.3 km2 of available hardbottom were covered with a minimum of 10% coral. Areas of moderately high to high coral cover (50% to 90% and 90% to 100% coral cover respectively) are found at depth and in sections of the park that are protected from high wave energy (Gibbs et al. 2007). Based on average coral cover from the categories outlined above, it was assumed that about 0.5 km2 of Kaloko was covered with 100% coral.  Based on the published literature, a wide variety of means exist to calculate coral biomass from coral cover (e.g., Odum and Odum 1955; Martinez-Estalella and Alcolado 1990; Crossland et al. 1991; McClanahan 1995). Calculated coral biomass for Kaloko varied markedly depending on the method applied, ranging from 10.342 t·km-2 to 8,000 t·km-2, reflecting the immense range (three orders of magnitude) in actual coral tissue biomass between different species (M. Hardt, Blue Ocean institute, pers. comm., 2008). Considering the estimates of Odum and Odum (1955) for Lobophyllia sp., biomass of heterotrophic tissue (polyps) and zooxanthellae in polyps were assumed to range between 0.021 g and 0.0038 g dry weight·cm-2 of coral skeleton. Based on coral cover calculated above, total coral biomass was estimated at 130 t·km-2. Values published in Atkinson and Grigg (1984) were used for dry to wet weight conversions. Porites spp., the dominant genus on the reef at Kaloko have massive hemispherical growth forms characterized by reproduction during a short period each year and slow growth. Given the paucity of information on population dynamics in most species of scleractinian the P/B was set to 0.14 year-1, based on information gleaned from Babcock (1991), who derived life history characteristics for 3 species of corals From the Tropics to the Poles, Wabnitz and Hoover  with relatively similar life characteristics to Porites spp.. This value is lower than those derived by Crossland et al. (1991) (1.095 year-1) or Arias Gonzalez et al. (1998) (21.68 year-1), and may in part reflect large differences recorded in the turnover time for different species of corals (Chadwick-Furman et al. 2000; Goffredo and Chadwick-Furman 2003). Consumption rates were found to vary widely between species, partly a reflection of location, depth, how much of a facultative consumer particular corals are, whether experiments were conducted in situ or in the laboratory, flow velocity, and the abundance of food particles. Hence, the Q/B of individual reefs is likely to differ markedly based on species composition. Corals possess symbiotic zooxanthellae in their polyp tissues (Odum and Odum 1955), and are therefore considered to be partly autotrophic. It was assumed here that autotrophic carbon contributed ~ 60% to animal respiration, based on findings from a study conducted on coral species that included Porites lobata and Porites compressa (Palardy et al. 2008), the most common species at Kaloko. Porites lobata feeding rates observed in situ (Johannes and Tepley 1974), combined with the biomass value, resulted in an estimate of Q/B equal to 5.84 year-1. However, this value was reduced to 2.1 year-1 based on the proportion of energy the species derives from heterotrophic feeding and evidence of patchy distribution of zooplankton over reefs (Palardy et al. 2006).  Coral diets included chiefly zooplankton, followed by detritus and phytoplankton (e.g., Anthony 1999; Rosenfeld et al. 1999; Ribes et al. 2003; Palardy et al. 2008). Octocoral Octocorals, represented here chiefly by Sarcothelia edmonsoni (also commonly spelled Sarcothelia edmondsoni), were included as a separate group as they are locally abundant (Beets et al. 2006; Marrack et al. 2009) and their sensitivity to water quality (Fabricius 2005) make them potentially important indicators of deteriorating conditions at Kaloko. Indeed, studies on the Great Barrier Reef suggest that Sarcothelia edmonsoni may be more strongly affected by declining water quality than hard corals (Fabricius 2005). Results from monitoring studies show significant variability in abundance between sites within Kaloko, with average cover estimated at 3.83% (range 0.7% ± 0.7% to 7.1% ± 5%) for surveys centered around 10 m depths (Marrack et al. 2009), and 10.1% in a study by Beets et al. (2010) that included shallower transects. Based on these findings, octocoral biomass was estimated to range between 1.62 t·km-2 and 4.22 t·km-2 and set at 2.9 t·km-2. Sarcothelia edmonsoni is a zooxanthellate octocoral, unlike most species of soft corals that do not contain symbiotic dinoflagellates in their tissue and therefore rely solely on heterotrophic nutrition. Despite benefiting from photosynthetic products produced by its endosymbionts, S. edmonsoni may derive a significant proportion of its carbon through suspension feeding of particulate and dissolved organic matter from the surrounding environment (Fabricius et al. 1995; Fabricius et al. 1998). It was therefore assumed that heterotrophically acquired carbon contributes about 80% of animal respiration. No studies have investigated the production or consumption rate of S. edmonsoni and published turnover rate values range widely depending on species (Chadwick-Furman et al. 2000). P/B was set to 0.2 year-1 based on a study by Goffredo and Lasker (2008), which calculated life history parameters for an octocoral as part of an adaptive management approach to a fishery on the species.  Data presented in Sorokin (1991) and Ribes et al. (2003a) led to a Q/B estimate of 9.25 year-1, which was reduced to 4.63 year-1 for the same reasons outlined for the functional group “corals” above.  Studies on tropical and Mediterranean gorgonians and soft corals have shown octocorals to capture detrital particulate organic matter (Ribes et al. 1999; Ribes et al. 2003a), small zooplankton prey (Coma et al. 1994; Rossi et al. 2004), and phytoplankton (Ribes et al. 1998; Picciano and Ferrier-Pages 2007). Evidence suggests that zooxanthellae-free soft corals in tropical reef environments are mainly herbivorous (Fabricius et al. 1995; Fabricius et al. 1998). In the absence of more detailed information, it was assumed that S. edmonsoni shows trophic opportunism with respect to sources of feeding and its diet included primarily phytoplankton, followed by detritus, and zooplankton (e.g., Orejas et al. 2003; Ribes et al. 2003; Tsounis et al. 2006). 84                  Baseline Trophic Relationships in Kaloko-Honokōkau, Hawai’i   Algae As turf (≤ 1 cm canopy height or frond extension) and macroalgae (> 1 cm canopy height or frond extension) are often used as indicators for nutrient availability and grazing pressure (Littler and Littler 1984), they were kept as separate functional groups. In addition, as green turtles predominantly feed on turf algae growing on the lava bench that spans the shallow subtidal/intertidal area, a separate functional group was created to represent their forage. Crustose coralline algae (CCA) were also included as a separate group. Surveys by Marrack et al. (2009) and Weijerman et al. (2009) showed that turf cover was equal to 33.13%, CCA cover 10.46%, and macroalgal cover 0.62%; corroborating findings by Beets et al. (2010). Turf samples from Kaloko-Honokōhau consisted of 20 different algal genera (McDermid et al. 2007). Other algae included Asparagopsis taxiformis, Caulerpa serrulata, Dictyota spp., Liagora sp., Sargassum sp., Turbinaria ornata, red gelatinous algae, and some geniculate corallines. Although no alien species were recorded by Marrack et al. (2009), the presence of Acanthophora spicifera has been documented on the west coast of Hawai‘i at three sites, including within Kaloko fishpond (C. Squair, University of Hawai‘i, pers. comm., 2008; Smith et al. 2002). Acanthophora spicifera has not been found on the reefs of Kaloko during dedicated surveys, or incidentally during coral cover surveys (S. Beavers, pers. Comm., 2008). No direct biomass estimates were made at Kaloko for any of the algae groups; but every effort was extended to ensure that values used were based on communities most similar to those found at Kaloko. Biomass estimates were derived based on data published by Smith et al. (2001) and those kindly provided by T. Sauvage (University of Hawai‘i, unpublished data) from experiments conducted in 2003 and 2004 in Waikiki; and estimates of percentage cover recorded within the park for each algae group (Beets et al. 2010; Marrack et al. 2009; Weijerman et al. 2009). Since intertidal turf communities are often much denser, a higher biomass estimate provided by J. Smith (unpublished data) was applied to the area encompassed by this habitat. Green turtles in Kaloko concentrate their foraging activities on turf algae growing on a shallow lava bench and intertidal area in Honokōhau Bay, the surface area of which was derived based on data presented in Gibbs et al. (2007) (0.026 km2). Calculations resulted in a biomass estimate for turf algae within Kaloko of 128.78 t·km-2 and 3.065 t·km-2 for turf algae restricted to the lava bench on which turtles feed. Macroalgae biomass at Kaloko was set to 22.691 t·km-2 and estimated based on data provided by T. Sauvage (unpublished data) from his experimental plots at Waikiki. CCA biomass was calculated based on Smith et al. (2001) and set at 37.818 t·km-2. The P/B ratio for turf algae was estimated at 19 year-1 based on Payri (2000) and falls within the range of values (12.5 year-1 - 30.8 year-1) provided by Polovina (1984), Klumpp and McKinnon (1992), Arias-González (1994), and Bozec et al. (2004) for coral reefs in the Pacific region. The productivity of turfLB was set higher (25 year-1) as shallow dense turf is known to register higher productivity rates than deeper reef turf (T. Sauvage, pers. comm., 2008). Naturally nutrient-rich groundwater that discharges in this area probably further contributes to elevated productivity rates. P/B for CCA and macroalgae were also calculated based on Payri (2000), and set at 1.777 year-1 and 9.824 year-1 respectively. Where necessary, conversion rates between gC, dry mass (DM) and wet weight (WW) were taken from Atkinson and Grigg (1984). Zooplankton The only data available for zooplankton biomass were taken from a station 150 m outside of Honokōhau Harbor (Bienfang 1980; Bienfang 1983). In all instances, samples were dominated by copepods (Bienfang 1983; Table 2 in Hoover and Gold (2005). Zooplankton biomass (g ww) for the entire park area was derived by integrating the mean tow biomass (44.96 mg m-3) over an average water column depth weighted by the surface area of each of 12 habitats, as listed in Gibbs et al. (2007) (27.58 m). This resulted in an estimate of 1.24 t·km-2. A review of the literature yielded a wide range of production rate estimates for zooplankton species, based in part on the ecosystem’s temperature, season, the zooplankton’s size fraction, diet, and the method used to infer production. The P/B was set to 219 year-1 based on the average of a range of production rates presented in Calbet et al. (2000). This value is comparable to the P/B value of 238 year-1, which was derived from production rates of the mesozooplankton fraction found at Uvéa atoll, New Caledonia (Le Borgne et al. 1997). Only the mesozooplankton fraction was used for comparative purposes given the 200 µm mesh size used for zooplankton sampling at Honokōhau (Bienfang 1980). From the Tropics to the Poles, Wabnitz and Hoover  Consumption rates for zooplankton vary depending on whether the zooplankton is mostly carnivorous or herbivorous with rates typically increasing with the herbivorous fraction of the zooplankton. Based on limited information, it was assumed here that copepods encountered in Kaloko are mostly herbivorous (Bienfang 1980; Bienfang 1983). Consumption rates presented in Calbet et al. (2000) were used to guide estimates and the Q/B was set at 949 year-1. Our estimate falls within the range of values found in or calculated from the literature: 269.29 year-1 as derived by Arias-González et al. (1997) for the Tiahura reef sector in French Polynesia; 550 year-1 based on estimates presented in Frost (1972); and 1,239.905 year-1 for the mesozooplankton fraction for an atoll community in Uvéa, whose taxonomic configuration is similar to that at Kaloko (Le Borgne et al. 1997). Phytoplankton Phytoplankton biomass values within park waters are considered to be low, ranging from ~ 0.14 mg chla  m-3 to 0.16 mg chla m-3 (Bienfang and Johnson 1980). Although the reason for the low biomass has been hypothesized to chiefly be attributable to control by zooplankton grazing (Oceanic Foundation 1975, in Hoover and Gold 2005), Bienfang and Johnson (1980) suggest that limited nutrient availability and other phenomena, such as hydrology, may be exerting stronger control over phytoplankton activity than zooplankton grazing. To estimate biomass the average value of measurements made in 1980 at 1.5 m and 5 m depth at the oceanic station 150 m outside of Honokōhau Harbor was used (0.1475 mg chla m-3). Values observed in the park between 1994 and 1996 and in 2000, at sites 100 m to 200 m offshore of Kaloko and ‘Aimakapa ponds (Brock and Kam 1997; Marine Consultants 2000 in Hoover and Gold 2005), fall within the range of the earlier studies. Following the same protocol as described above for zooplankton, phytoplankton biomass was set at 3.295 t·km-2. This value is identical to that derived by Polovina (1980) for French Frigate Shoals, but substantially higher than that calculated by Arias-González et al. (1997) for Moorea, French Polynesia (0.32 t·km-2).  Phytoplankton productivity rates were set at 325.458 year-1, based on measurements made by Bienfang and Johnson (1980) outside the harbor. This value is somewhat lower than the P/B of 475 year-1 derived for the Great Barrier Reef (Furnas et al. 1990). For Uvéa atoll, New Caledonia, Le Borgne (1997) estimated P/B to range between 716 year-1 and 511 year-1, depending on the C: Chl a ratio used. C: Chl a ratios in the literature are highly variable, e.g., 84 (Charpy and Blanchot 1998), 60 (Yahel et al. 1998; Barbosa et al. 2001), 20 to 160 (Taylor et al. 1997). It is likely that a higher ratio is more appropriate for Kaloko, with values as high as 200 applicable to open ocean oligotrophic environments (Gasol et al. 1997), and was set to 90.  All basic input parameters for the 26 groups included in the model are presented in table 1.  Model analysis, indices, and uncertainty around input data  Total trophic flows within the ecosystem in terms of consumption, production, respiration, exports and imports, and flow to detritus (t·km-2·year-1) were quantified. Ecological indices as obtained through Ecological Network Analysis (ENA) were then used to describe the ecosystem structure. ENA is a modeling technique used for understanding the structure and flow of material between components of an ecosystem, as modeled (Ulanowicz 1997). It is descriptive in nature and is most commonly used for evaluating food webs (Wulff et al. 1989; Christensen and Pauly 1993). It is integrated into Ecopath and allows for the calculation of ecosystem macro-descriptors, which quantify trophic structure, organic matter recycling, and ecosystem size and organization. These descriptors include total system throughput (T), Ascendency (A), Development Capacity (C), and the relative overhead (O/C). Throughput describes the size of a system and represents a measure of its metabolism (Christensen and Pauly 1993). Ascendency integrates both size and organization of a given system (Christensen 1995). The Development Capacity quantifies the upper limit to Ascendency whilst the system’s overhead (O) is complementary to the Ascendency and measures to which degree particular links can be considered ‘redundant’ (Heymans 2003). The O/C ratio was proposed by Heymans (2003) as an index of the resilience of the system (i.e., it can be seen as an index of the system’s ability to withstand disturbance (Ulanowicz 1997)). ENA further allows the derivation of Transfer Efficiencies (TE), summarizing the proportion of consumption that is passed up a food web. The TE is obtained by calculating the ratio between the production of a given trophic level and the preceding trophic level (Pauly and Christensen 1995). Finally, the mixed trophic impact (MTI) analysis (Ulanowicz and Puccia 1990), or ecological input-output, was run. MTI describes how any functional group impacts, directly and indirectly (i.e., both predatory and competitive interactions), all other functional groups within the food network. In other words, this analysis provides a first-order quantification of the beneficial 86                  Baseline Trophic Relationships in Kaloko-Honokōkau, Hawai’i   and negative impacts of one group on another (scaled between -1 and 1). The MTI for living groups is calculated by constructing an n x n matrix, where the interactions between the impacting group (j) and the impacted group (i) is represented as follows (Christensen and Walters 2004): ܯܶܫ௝௜ ൌ ܦܥ௝௜ െ ܨܥ௜௝ where DCji as previously defined, is the diet composition term expressing how much (i) contributes to the diet of predator (j), and FCij is a host composition term giving the proportion of the predation on (i) that is due to (j). For detritus groups, the DCji term in the MTI analysis is set to 0. Addressing Uncertainty Any ecosystem model realisation requires acknowledging the large amount of data required in its development, and the difficulty in quantifying the flows between the food web’s individual functional groups. Functional group dynamics can be verified by fitting model data to actual population trends over time. Such time series data do not exist for Kaloko. To address model uncertainties, probability distributions for all Ecopath input parameters (including the diet compositions matrices) were entered through the ‘pedigree’ function (Funtowicz and Ravetz 1990) of Ecopath (Pauly et al. 2000). Using a Monte Carlo re-sampling routine, the ‘Ecoranger’ module of Ecopath draws random input variables within the confidence intervals defined for each parameter type in the pedigree tables and uses these as prior probability distributions for all input data. This approach leads to a large number of model realizations that are evaluated for their conformity to user-defined criteria as well as physiological and mass-balance constraints. The results include probability distributions for the estimated parameters along with distributions of parameters in the accepted model realizations. This routine can be run for the model overall, as pedigrees were associated with all input parameters. However, given the primary goal of this study to determine whether green turtles are at carrying capacity at Kaloko, the focus was placed on sources of uncertainty associated with: estimates of green turtle biomass, P/B and Q/B; the proportion of sea urchins’ diet that was derived from turf algae growing on the nearshore lava bench (turfLB); and variability associated with turfLB biomass and P/B. RESULTS Trophic parameters for the 26 groups of the final balanced model for Kaloko are presented in table 1 (outputs are in bold). Trophic flows between all functional groups are depicted in figure 2. Sea urchins (Tripneustes gratilla, Echinometra mathaei, Heterocentrotus mammilatus, and Echinothrix spp.) accounted for the largest proportion of total living biomass in the system (40%) (figure 3). Green turtles and reef fish groups, including ‘sharks and jacks’, only represented 0.2% and 5.5% of total biomass respectively (figure 3). Reef fish functional groups were dominated by herbivorous and mobile invertebrate feeding species. Not surprisingly, consumption by sea urchins had the biggest impact (45%) on available resources at Kaloko; whereas fish accounted for 14.4% and green turtles 0.2% of total consumption. Total fisheries catches represented less than 1 % of the total fish biomass (table 1). ‘Sharks and jacks’ were caught in larger quantity compared with their relatively low biomass in the assemblage. Mean trophic level of the total fisheries catch was 2.59 (table 2). Ecotrophic efficiency (EE) values (the proportion of the production used within the system) were lowest for some of the highest trophic levels, including ‘spinner dolphins’, ‘monk seals’, and ‘sea birds’, as well as for ‘crown of thorns starfish’, ‘hawksbill turtles’, ‘green turtles’, ‘urchins’, and ‘corals’. For the first three groups, this valuation is due to these species deriving a significant proportion of their food from outside park waters. They were included in the model chiefly for purposes of representation (i.e., to acknowledge that these species occur in the park, and may at some point in the future suffer from indirect effects of park development, even if they do not feed exclusively or primarily in park waters). For the five other groups, the low EE values are a result of low predation and fishing pressure being exerted on these species. Fish groups, overall, also had relatively low EE values, with higher values registered by those species pools that were subject to higher fishing mortality. This valuation suggests that the system generates a reasonable amount of surplus secondary production. Indeed, the largest component of the mortality coefficients within the system at Kaloko was due to predation mortalities, with the exception of ‘sharks and jacks’ for which the fishing mortality rate was greater than predation mortality explained within the system. In contrast, most of the production by the macro- and turf algal groups, ‘zooplankton’, ‘benthic invertebrates’, and ‘phytoplankton’ was accounted for through consumption by other trophic groups From the Tropics to the Poles, Wabnitz and Hoover  within the model (EE > 90%). In other words, urchins, herbivorous fish, and green turtles maintained all algae at Kaloko in a cropped state. Reef building corals, octocorals, benthic invertebrates and zooplanktivorous fish were the main predators of zooplankton, whilst zooplankton consumed most of the phytoplankton production in the system. Benthic invertebrates were predated upon mostly by MIF reef fish and species within the benthic invertebrate group itself.   Throughput (total flows) values for each functional group highlighted the importance of turf algae, zooplankton, phytoplankton and sea urchins in system structure. The high EEs attained for all algal groups indicated that the system was at carrying capacity with respect to grazing, including for green turtles. The primary producers’ high EEs were also reflected in the low production/respiration ratio of 1.12 for the system overall (table 2). Focusing on green turtles, urchins, and herbivorous reef fish in the system only, the Mixed Trophic Impact (MTI) routine underlined the competition for resources both within and among the three groups (figure 4). Sea urchins were responsible for the largest among-group effect, impacting negatively on both herbivorous reef fish and green turtles, and overall had the largest impact on algal and detritus resources (not shown). Given green turtles’ feeding preference on turfLB, they were the group most impacted by changes in turf algae. When accounting for the uncertainty around parameters affecting productivity and consumption of turfLB, EE values of turfLB in successfully balanced model runs ranged between 0.416 and 0.998.  The mean Transfer efficiency (TE) in the ecosystem as a whole was 4.6%, with a value of 4.5% for flows originating from primary producers and 4.7% from the detritus. This was primarily due to poor transfer efficiencies at the higher trophic levels. This low efficiency was underlined by consumption dominating total system throughput for the lower trophic levels whereas respiration and flow to detritus dominated Green sea turtlesReef fish - herbivoresAll other reef fish (incl. rays)UrchinsCoralsAlgaePlanktonOther invertebratesMammals, hawksbill sea turtles and birdsFigure 3: Proportions (%), in terms of biomass (t·km-2) of aggregated functional groups at Kaloko. Sum of all consumption 5,332.03             t/km²/yearSum of all exports 520.07                t/km²/yearSum of all respiratory flows 3,477.31             t/km²/yearSum of all flows into detritus 1,700.15             t/km²/yearTotal system throughput 11,030.00           t/km²/yearCalculated total net primary production 3,895.09             t/km²/yearTotal primary production/total respiration 1.12                    Total biomass (excluding detritus) 699.53                t/km²Total catches 0.35                    t/km²/yearMean trophic level of the catch 2.59                    Throughput cycled (excluding detritus) 54.52 t/km²/yearFinn's cycling index 6.13 % of total throughputAscendency 31.50                  %Relative overhead (O/C) 68.50                  %Table 2:  Summary of outputs from the Ecological Network Analysis 88                  Baseline Trophic Relationships in Kaloko-Honokōkau, Hawai’i   the higher trophic levels in the model (figure 5). This result would be expected in a system with low biomasses at the highest trophic levels. Some upper trophic level species such as trevally (also locally known as ‘ulua’ or ‘jacks’) are highly vagile. Thus, although they might not be heavily fished directly in park waters, their abundance may still be depressed due to high fishing pressure on this group along the remainder of the Kona Coast. Ascendancy was estimated at 31.5% of capacity and relative overhead was 68.5% (table 2), 46% of which was attributable to internal flows indicating that the system at Kaloko contains a number of ‘redundant’ trophic linkages. These observations are consistent with a system exhibiting relatively high resilience to perturbation with respect to energy flows, or a high system stability sensu Odum (1971).  Net primary productivity was 3,895 t·km-2·year-1 and was accompanied by a high flow to the detritus pool (table 2). Turf algae (including turfLB) and phytoplankton accounted for 57% and 24% of total production in the system, or 65% and 27% of total primary production respectively. Sea urchins (38%) were the major contributors to the detrital pool, followed by by zooplankton, while the combined fish groups (mainly herbivores) contributed 11% of all flows to the detritus. The ratio between primary production and respiration (PP/R) indicates that the system is at a low developmental stage sensu Odum (1969; 1971). This finding was corroborated by the low Finn’s cycling index, showing that only a small fraction of the throughput (including detritus) gets recycled (6%; table 2). The proportion of flows originating from the detritus was 0.27%. The overall pedigree index of the model (0.539) was relatively high in comparison to 50 previously constructed models for which pedigree values ranged between 0.164 and 0.676 (Morissette 2007). -0.4-0.3-0.2-0.100.10.2Green sea turtlesReef fishes -herbivoresUrchinsMacroalgaeTurf algaeTurf algae_lavabenchDetritusImpacted functional groupsMTIGreen sea turtlesReef fishes - herbivoresUrchinsTurf algaeTurf algae_lava benchMacroalgaeImpacting functional groupsFigure 4: Mixed trophic impact analysis for herbivores at KAHO. Impacts of increases in the biomass of a particular group (impacting) on another (impacted), resulting in an increase in the latter’s biomass, are recorded as a positive on the y-axis. Impacts resulting in a decline of the impacted group are recorded as a negative value. Group names along the x-axis represent the impacting groups, whilst the legend provides color codes for the impacted groups. From the Tropics to the Poles, Wabnitz and Hoover  DISCUSSION Description of the Kaloko-Honokōhau System Quantitative descriptions of the flux of matter and energy can provide significant insights into the fundamental structure of ecosystems. The Kaloko system is dominated by primary production (PP), 27% of which is contributed by phytoplankton and 65% by algal turfs, which is slightly lower than other published estimates for primary producers in tropical systems (e.g., Wanders 1976; Adey and Steneck 1985; Adey and Gotmiller 1987). The trophic networks at Kaloko were dominated by grazing, with herbivores accounting for 43% of all living biomass within the system (figure 3), of which 93% was sea urchins. This grazer dominance was further highlighted by the high ecotrophic efficiencies achieved for the main PP functional groups, including phytoplankton. Results from rapid assessment surveys throughout the state confirm that sea urchins in particular, and herbivorous fish, are common on reefs in Hawai‘i (Rodgers et al. 2004). In contrast, relatively low EEs of higher trophic functional groups indicate that the foraging activities of herbivores are not limited by predation pressure, as demonstrated by our analysis of total system throughput (figure 5). Indeed, few predators commonly occur in the park, and fewer still have been observed feeding in the park. Therefore, as highlighted by the model’s average TE of 4.6%, much lower than the mean of 10% obtained for various other ecosystems (Christensen and Pauly 1993), only a small proportion of production is being transferred up the food chain. Comparison between two sets of underwater visual census data (E. Brown and NPS, unpublished data) showed a 5-fold increase in piscivore biomass between 2005 and 2007. This inter-annual differences may have been due to real differences between the two datasets, possibly reflecting the restrictions on gillnetting in park waters implemented in August 2005. Differences in species biomass and composition may also reflect either variability in survey accuracy due to some species displaying vagile (e.g., Caranx melampygus) or cryptic behavior (e.g., Gymnothorax flavimarginatus, Gymnothorax meleagris, Cephalopholis argus), or seasonal changes (Friedlander and Parrish 1998; Vitousek et al. 2009) as the 2005 and 2007 datasets were collected in April and October respectively. Future surveys should therefore focus on good intra-annual coverage to provide data that are representative of, and integrate, seasonal variation.  Phytoplankton and zooplankton had some of the highest EEs. The waters around Hawai‘i are generally oligotrophic (Bienfang et al. 2009), and consequently low biomass of phytoplankton and zooplankton groups is to be expected. Reef environments along the Kona coast are close to the deep slope of the Pacific and subject to strong wave action and currents along the shore (Presto et al. 2007). Therefore, it was assumed that zooplanktivores on the reef actually derive a substantial portion of their diet from open-ocean plankton. A further assumption was made: that ocean plankton also contributes to the energy intake of benthic invertebrate feeders and other functional groups that rely on plankton for a portion of their diet. Experiments should be conducted to ascertain, quantitatively, the proportion of ocean plankton in the diet of reef organisms at Kaloko, or other similar systems in Hawai‘i.   0%20%40%60%80%100%I II III IVTrophic levelstotal system throughput (%)ABCDFigure 5: Fate of total system throughput (A = Respiration, B = Flow to detritus, C = Export, D = Consumption by predator) in percentage-per-integer trophic level. 90                  Baseline Trophic Relationships in Kaloko-Honokōkau, Hawai’i   At 42.54 t·km-2, the model’s benthic invertberate (excluding sea urchins) biomass was similar to estimates provided for models in Raja Ampat, Indonesia (51.68 t·km-2; Ainsworth et al. 2007), and GBR, Australia (61.41 t·km-2; Tudman 2001), but substantially lower than biomass estimates for the Moorea barrier reef and fringing reefs, French Polynesia (B = 198.26 t·km-2 (Arias-González 1994) and 322.8 t·km-2 (Arias-González et al. 1997) respectively). Few studies have investigated the composition of benthic invertebrates on reefs, yet they are an important prey for certain groups of reef fish (e.g., Hobson 1974), and can represent an important fishery (Friedlander and Parrish 1997). Given that benthic invertebrates accounted for 6% of total system biomass, species composition and their relative contribution to total group biomass of the invertebrate community at Kaloko should be determined. Sea urchin densities at Kaloko (~ 5 individual per m-2 for E. mathaei and < 1 individual per m-2 for all other species) were comparable to sea urchin densities recorded on reefs throughout the state of Hawai‘i (Rodgers et al. 2004) and similar to, or lower than, those reported from other locations (e.g., up to 5 to 6.8 individuals per m-2 (chiefly E. mathaei and T. gratilla) at La Réunion (Naim et al. 1997) and 2 to 4 individuals per m-2 (Echinometra sp.) in Fiji (Appana and Vuki 2006). Current sea urchin densities at Kaloko are also similar to those recorded in a study conducted in the 1970s (Ebert 1971). These similarities suggest that high urchin densities are natural and do not represent a release from predation pressure due to increased fishing pressure in the last 30 years, as noted in a number of locations along the coast of Kenya (Muthiga and McClanahan 1987; McClanahan 1998). Given sea urchins’ primary role in maintaining algae closely cropped on the reef and therefore their contribution to ecosystem resilience, future survey efforts should focus on obtaining more detailed data for urchin species, including:  1. Biomass for all reef zones - It is important to note that the sea urchin biomass was derived from survey data collected at randomly selected sites of relatively high coral cover at 10 m depth (Marrack et al. 2009; Weijerman et al. 2009), thought to be representative of sea urchin abundances throughout the park on similar hard bottom composition (M. Weijerman pers. comm., 2008). Complementary macroinvertebrate-focused surveys recording test size for individuals of the different sea urchin species recorded along all transects, with surveys conducted at all depths, would considerably help refine values used here. Different sized sea urchins from all species should be selected and a species-specific length-mass relationship established. Surveys should also ensure to cover both subtidal and intertidal zones as urchin community structure can differ substantially between them (Ebert 1971). For example, the shingle urchin (Colobocentrotus atratus) was recorded as one of the most common herbivores in the intertidal habitat in 1990 (Parrish et al. 1990). No data for this species were available for inclusion here as they were not targeted by recent surveys. Moreover, other studies have reported sea urchins to be patchily distributed and therefore, the averages used here and extrapolated to the entire hardbottom reef area (with ≤10% coral cover) may overestimate (or underestimate) actual sea urchin abundance; 2. P/B in the wild; 3. Q/B in the wild – most of the available data on consumption rates are confounded by the use of different methods for estimating (i) gut turnover (some echinoids feed only during the night, whilst other species eat continuously through 24h); and (ii) feeding rates in the field versus laboratory conditions where individuals are fed ad libitum. Additional variation in published information may stem from how individual researchers measure wet weight biomass of different urchin species. Depending on time of the day and state of gut evacuation, for example, mass of individual sea urchins may vary considerably which in turn may impact consumption rate calculations (Carreiro-Silva and McClanahan 2001); and 4. Dietary preferences in the different habitat zones based on abundance of food items – Surveys on reefs in Hawai‘i indicate that certain algae species seem to be favored over others (Birkeland 1989; Stimson et al. 2007). In particular, attention should be paid to the proportion of turf algae to macroalgae consumed by sea urchins. Surveys should be conducted seasonally and include night-dives as published evidence indicates that echinoid species differ in foraging behaviors, with some feeding both day and night, while others are nocturnal (Vaitilingon et al. 2003). Note that some species are known to fill a large proportion of their gut with reef framework material (CaCo3) (e.g., Echinothrix sp. and E. mathaei) (Bak 1990; Mills et al. 2000; Carreiro-Silva and McClanahan From the Tropics to the Poles, Wabnitz and Hoover  2001). However, as sea urchins do not derive energy from this material, this impact is not considered under the current modeling scenario.  Turtles at Carrying Capacity Green turtles are at carrying capacity at Kaloko - based on (i) their biomass estimates and consumption rates, (ii) estimates of turfLB as well as the algae’s primary production rates, and (iii) the high degree to which sea urchins feed on the turtles’ main food resour